QSVM_scratch.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import os\n",
    "import subprocess\n",
    "import time\n",
    "import pickle\n",
    "import numpy as np\n",
    "import pandas as pd\n",
    "import matplotlib as mpl\n",
    "from matplotlib import pyplot as plt\n",
    "from sklearn import preprocessing\n",
    "from tqdm.notebook import tqdm  #　for jupyter\n",
    "# from tqdm import tqdm  # for else\n",
    "\n",
    "from qiskit import QuantumRegister,ClassicalRegister,QuantumCircuit, execute\n",
    "from qiskit.visualization import plot_histogram\n",
    "\n",
    "\n",
    "from qiskit import BasicAer\n",
    "from qiskit.circuit.library import ZZFeatureMap\n",
    "from qiskit.ml.datasets import ad_hoc_data, breast_cancer\n",
    "from qiskit.aqua import aqua_globals, QuantumInstance\n",
    "from qiskit.aqua.utils import split_dataset_to_data_and_labels, map_label_to_class_name\n",
    "from qiskit.aqua.algorithms import SklearnSVM\n",
    "from qiskit.aqua.algorithms import QSVM\n",
    "\n",
    "from scipy.linalg import norm\n",
    "import cvxopt\n",
    "import cvxopt.solvers\n",
    "from pylab import linspace, scatter, meshgrid, contour, array\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import seaborn as sns\n",
    "\n",
    "\n",
    "\n",
    "seed = 10598\n",
    "backend = BasicAer.get_backend('qasm_simulator')\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "alpha = Parameter(\"α\")\n",
    "beta = Parameter(\"β\")\n",
    "gamma = Parameter(\"γ\")\n",
    "delta = Parameter(\"δ\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 447.652x144.48 with 1 Axes>"
      ]
     },
     "execution_count": 13,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Implementation from Library\n",
    "data = [alpha, beta]\n",
    "ZZFeatureMap(feature_dimension=2, reps=1).assign_parameters(data).draw(output=\"mpl\")\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Implementation from Scratch\n",
    "def feature_map_zz_dim2(data,rep=2):\n",
    "    q = QuantumRegister(2)\n",
    "    c = ClassicalRegister(2)\n",
    "    qc = QuantumCircuit(q,c)\n",
    "    for i in range(rep):\n",
    "        qc.h(q)\n",
    "        # exp(i*a*Z_1)\n",
    "        qc.rz(-2*data[0],q[0])\n",
    "        # exp(i*b*Z_2)\n",
    "        qc.rz(-2*data[1],q[1])\n",
    "        # exp(i*(π-a)(π-b) Z_1*Z_2)\n",
    "        qc.cx(0,1)\n",
    "        qc.rz(-2*(np.pi-data[0])*(np.pi-data[1]),q[1])\n",
    "        qc.cx(0,1)\n",
    "    return qc"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXQAAACoCAYAAADw6BWzAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAZ6UlEQVR4nO3de1yUdd7/8deAB0DAA5gHVFTEQ4poklmWYGmp9dvdTC0Pbf70p9zWdm7bu9DHtppaZtu9rXu7B7efbmWZ2K3tRgesINPKwEzJFA8QIEgqyskjzNx/zIogoDM41xwu3s/HYx4Pva7vXN/PXNfMey6+12EsNpvNhoiI+Dw/TxcgIiKuoUAXETEJBbqIiEko0EVETEKBLiJiEgp0ERGTUKCLiJiEAl1ExCQU6CIiJqFAFxExCQW6iIhJKNBFRExCgS4iYhIKdBERk1Cgi4iYhAJdRMQkFOgiIiahQBcRMQkFuoiISSjQRURMQoEuImISCnQREZNQoIuImIQCXUTEJBToIiImoUAXETGJFp4uoLl5NwMOn/BM3xHtYWKc88/b9ymU/+T6ehwRcg30u9UzfYv4GgW6mx0+AQc9FI5NVf4TnCzwdBUiciUachERMQkFuoiISSjQRURMQoEuImISOigqLvPkygR++PFL/P1b4ufnT+f2vZh2WxLxsZM9XZpIs6BAF5eaPmYB08fMp7q6ik3bVrB07TT6RAwlIryPp0sTMT0NuYgh/P1bMP6GOVRbqzhYuNPT5Yg0Cz4d6AUFBTzyyCPcdNNNBAUFYbFYyMrKanI7cZ3zVef417aVAHQL7+vhakSaB58ecjlw4ADr1q1j2LBhxMfH8+GHH15VO7l6az9ZzPr05Zw+W46/f0uemLyK3l0HA/DB9r+zOfP1mrZFJYeI6XULz0x701Pl1lN5FnbkwolKCGgJsT2gU1tPVyXewmqFPYWQe8z+/57hcG1X8POSXWOfDvRRo0ZRXFwMwOrVqxsNakfbeaPk5xPoMWgMw38x36HpnjbttiSmj5lP+akTvLx+NjsPfMr44bMBGD98ds2/S8qO8NRfRjNz3POeLLeGzQapWfBRFlRbL05P2QUDI2DGTRDYynP1ieflHIU1X8DJU3WntwuCB26GXh09U1dtXvK9Up/VamX58uVER0cTEBBAbGws6enp9OvXj7lz5wLg5+DXoqPtxHVCgtrzxORVbN+bwrasTXXmWa1Wlr41nVnjltClQy8PVVjX5u/t4V07zC/4/jD8Na3hedI8FJ6A//4ESk/Vn1d6yj6v0EP3aKrNa5Nu1qxZLFq0iMTERD744AOmTJnC1KlTOXToEMOGDfN0eeKA0KAO3HPLE7z24bNYrRfT8PXU39Grcww3x9ztweouOnUWPtp9+TY5R+3BLs3TR7uhqhpsDcyzYZ/3kRcclvPKQF+7di1r1qzhvffe46mnnmL06NEkJSVx4403UlVVZWigFxcXc/vttxMUFERsbCzffvutYX01B3ff8iglZUWkZv4DgB37PyEz+2Pm3LnMw5Vd9O2PUHWFvW+LBb464J56xLtUnoVdBQ2H+QU2YHe+va0neWWgL126lHHjxhEfH19nep8+fWjZsiUxMTGG9T1v3jz69+/P8ePHeeihh5g0aRLV1dVXfJ7FYnHokZ6e5nRN2zctZuXcdnUehdlfOL2c9PQ0h+tsSs0vz0tj+pi6Y/ptAkJ5d2EJd1w/k5KyI6zY+Cuenf4WLVs4NiDd1Jqdefx6/mJstst9XO1j7J9/vcvwWvTwvkePqIFc4e0BgNUGPaIGurx/Z3jdQdGCggKysrJ4/PHH683Ly8tj4MCBtG7d2pC+y8vLef/99zl8+DCBgYHMnTuXJUuW8NVXXzFy5EhD+nTE8J8nNXhQ1Ne8sXkRlWdKeWndzJpp3Tv247FJf/FcUcC5U6VX/ODYrNWcrTzpporEm5w7XWpIWyN4ZaADdO7cuc7006dPk56ezoQJEwzre//+/YSFhREeHl4zLSYmhj179lwx0K+0h3fBH1M9dz/0+PgEkp93rM7aMt52zf3QH5n4Jx6Z+CennhMfn4BtpfM1O+NYOTz/3uXbWPz8eWTGKNYvMrYW8U6vfAh5xxsfdrFYoEcYlB/37A8HeN2Qy4Uwzc7OrjN92bJlFBUVcd111xnWd2VlJaGhoXWmhYaGUlFRYVif4nnhITCkR+PzLRYICYDre7uvJvEuYwddYQzdBmMHuq2cRnndHnrv3r0ZPHgwS5YsoUOHDkRERJCcnExKSgpAvQOiycnJAGRkZACQmprK3r17adOmDePHj3eqXZs2bSgvL6+z/LKyMoKDgw14peJNpo6AU+cg+4g9wGv/wRXcGh68zX6hkTRPg7rZf77x3YyG50+Ms7fxNIvN0bECN8rOziYxMZHt27cTFhbGAw88QEhICElJSZSVlREYGFjTtrGxz8jISHJzc51qV15eTnh4OIWFhYSFhQHQq1cv3njjDZeNoXtyyCXqGnh4rPPPc9WQS1O06wZx97mnL6sN9hbCVwdhV7592uTrYVgvhbnYHS2DL/ZD+l77/+P7w83R0DH08s9zF6/bQwfo27cvn332WZ1p999/PwMGDKgT5uD42LUj7UJCQrjzzjtZtGgRL7zwAm+88QYWi4URI0Y4Xrz4LD8LXBthfzz277sRjNRtaKSWjqFw97CLgX63l10S45WB3pCMjAy3BOvKlSuZMWMG7du3Jzo6mg0bNuDv7294v2byQ97X/Pm9x/Hz86dvtzjm/ewVKk6frLnrYlTXIQQHtvNwlSLm4xOBXlFRQXZ2Ng8++KDhfXXq1InU1FTD+zGzTu0ieSnxU1q1DGDp2unkFO2m+MSPbM36H2zYqDxTxk0Df+bpMkVMxycCPTg42KGLe8Q7dAi9eMqpv18L/Pzsf+HYap0ncK7qLEvfnMbR0nxuj5vJ/oJMnpzyd7fXKmImPhHo4psOFe6itPIYkZ2uJSy0K4Gt7WcLRXUdwhe7NxDdbRjz73+HmS9G88uxz3m2WBETUKBLk5WUHWHxm3VPQekQ0pmkGW9TdqqEFRt/xfwZ7wAQHNiO2KiEmnbFJ34kqusQ/P386dlpIHH97nBn6SKmpECXJusQ2pmX56XVm15dXcULb81gzl0v1Rl+qS28bTeOlxVitVrJPZJFVfV5g6sVMT+vu1JUfF/6rvVk53/Dqvd/w5MrE9iT+2W9NjfHTOTbA5+wZO1Ubo+byd/ef9oDlYqYi/bQxeVuHTqVW4dOvWybwFZtSJr+lpsqEmkeFOhuFtHe9/oOuca1dfhK3yK+RoHuZhPjPF2B8/rd6ukKRMQRGkMXETEJBbqIiEko0EVETEKBLiJiEgp0ERGTUKCLiJiEAl1ExCQU6CIiJqFAFxExCQW6iIhJKNBFRExCgS4iYhIKdBERk9DdFt3s3Qw4fMIzfUe0b9rdHvd9CuU/ub4eR4Rc07zu9ujJde2rmtt75HIU6G52+AQc9LEPbPlPcLLA01U0D1rXcjU05CIiYhIKdBERk1Cgi9Ryrgryjl/8f/kZz9Ui3sdqg+Kyi/8vLrNP8xYaQ5dmr/IsbD8E3+RA0Umw1fqALtgA7YIgpjvcHA2d2nquTvEMqxX2FMK2/fbjX2erLs5b+k9o3QKiroGbouHaruDnwd1kBbq4zJMrE/jhxy/x92+Jn58/ndv3YtptScTHTvZ0aQ2y2mBrNvxzp33PvDEnT8GWffbHDVHwi+sgsJX76nQnX9uGRss/Dmu/hKLSxtucrbIH/p5C6NIOpo2A7mHuq7E2Bbq41PQxC5g+Zj7V1VVs2raCpWun0SdiKBHhfTxdWh1nz8P/3wJ7i5x73tcHYV8RJI62f3jNyFe2odE+3wcbM50bUik6Ca98BHcPg1v6GVdbYzSGLobw92/B+BvmUG2t4mDhTk+XU8e5KvhrmvNhfsHJU/DHVDhymb02M/DmbWi09L32a0aaMj5utcGGDPsy3E2BLoY4X3WOf21bCUC38L4erqau97+78rUA/zXd/mjMqXOwZgtUVbu2Nm/izdvQSLnH7Hvml3Ol9wfYl5F7zHV1OcKnh1wKCgpYtmwZGRkZ7Ny5k9OnT7N7924GDRrUpHZy9dZ+spj16cs5fbYcf/+WPDF5Fb27Dgbgg+1/Z3Pm6zVti0oOEdPrFp6Z9qbb6ss5Cp+7aM+pqBQ+zoIJsa5Znrfw9m1opKpq+5i5K05csWFf1tMToIW/CxboAJ/eQz9w4ADr1q2jXbt2xMfHX3U7uXrTbkti46KTJD93jOH9J7DzwKc188YPn83L89J4eV4aSdPfJqBVG2aOe96t9W3+3jUf1gs+32cfjzcTb9+GRtqVDz+VXbmdo34qsy/TXXw60EeNGkVxcTEpKSnce++9V93OGyU/n8D2jfU/MI1N9xYhQe15YvIqtu9NYVvWpjrzrFYrS9+azqxxS+jSoZfbajpRCXsOu3aZZ85DZq5rl+ktvHEbGm3rft9YZmO8NtCtVivLly8nOjqagIAAYmNjSU9Pp1+/fsydOxcAPwdP+HS0nbhWaFAH7rnlCV778FmsVmvN9NdTf0evzjHcHHO3W+vZX+zavfMLso8YsFAv4W3b0EjnquxDcq6Wc/Typ8W6ktcm3axZs1i0aBGJiYl88MEHTJkyhalTp3Lo0CGGDRvm6fLEQXff8iglZUWkZv4DgB37PyEz+2Pm3LnM7bXkl/jWcr2FN21DIxWeNOaqT6vNvmx38MqDomvXrmXNmjWkpaXVjHmPHj2aHTt28O677xoa6L/97W9Zv349e/fu5Z133mHSpEkOPc9isTjU7p6kz+g2IOEqKmy69PQ0Hrl9tNPPW/4fnxEblXDFdi/PS6s3rU1AKO8utCdeSdkRVmz8FUtmf0DLFo5dmZOensb1U52vuSF3PrqBPtdPrDPtSmcqNDb/sVrHAItLzmCxBF5ldXaOrmujGLENjeaq90if6ydy56Mb6kxzxfsDYMz4iRzM+J8m1WWzOf4t45WBvnTpUsaNG1fvAGafPn1o2bIlMTExhvUdHR3NH/7wBxYsWGBYH87avmkxmSnL60w7f6aCHoPGeKiipnlj8yIqz5Ty0rqZNdO6d+zHY5P+4p4CHPzS9ZrleiGPb0NDGbcdHd3hu1peF+gFBQVkZWXx+OOP15uXl5fHwIEDad26tWH9z5gxA4DFixc79TxHv0X/mOr8/dCH/zyJ4b+YX2da8vMJzi0EiI9PIPl55/+mzHjbNffofmTin3hk4p+cek58fAK2la75O3jd1/DlgbrTLt2TuuDCnldj82sLC23t1F7U5bhqXRulKdvQaK56jxwohhWb605zxfsD4MN/biCqU9Nrc5TXjaEXFNjfzZ07d64z/fTp06Snp2v8XJqsWwdjluup+3aIa0W0N3DZBr33LuV1gR4eHg5AdnZ2nenLli2jqKiI6667zhNliQlEXWPMcnt3NGa54l6BrYwJ9Yj2ENDS9cttiNcNufTu3ZvBgwezZMkSOnToQEREBMnJyaSkpADU20NPTk4GICMjA4DU1FT27t1LmzZtGD9+vNPtxLw6t7WH7yEXnprm7wfDe7tueeJZN/aB5G9cu8yb3HhPM68LdD8/P9avX09iYiLz5s0jLCyMBx54gIceeoikpCQGDx5cp/3kyXVv6/nEE08AEBkZSW5urtPtvM2k+WlOTfcGP+R9zZ/fexw/P3/6dotj3s9eoeL0yZobPEV1HUKbgLYs/MckjpUdpnvHfjx93xq31DZ6gGsDfXhvCA5w3fJqW5+2nKHRt9EnYuhVLcdms9Vb1wcLvyMz+2OmJPy63rYJDmxXp29Htmdw4NXferKhfhpSu3ZXi+sFH+123Q+bhATAMDded+V1Qy4Affv25bPPPqOyspK8vDwWLVrE7t27GTBgAIGBdU8Ps9lsDT4uDWlH250/f54zZ85gtVpr/u2qA17NRad2kbyU+CmvPLiFkxU/kVO0m6ycL9ic+Tqpmf9g16HPySnaTVBAKH98+CtKK49RUuaeq3NiusOQHq5ZVttA+D9Xl7WNslqtfJ+79arDHGhwXUd1jWVP7jasVmu9bXNp345sT1doqJ+G1K7d1QJawuThrlvelOHuG24BL9xDb0xGRgYjRowwvJ85c+awZo19b3HLli0A5OTk0LNnT8P7NosOoRcPaPv7tcDPz35nIlut6zTLT58gvG0E586f4WTFT4QEuemoETDpejh8Ao6WN97mSmcvtPCDGSMhyKBTsQ8VfUd4u24AfHcwjefW3E3vLrEcKcmhT8RQfjdzo8PLamxdR4RH1+xl1942tfsGx7bnuaqzLH1zGkdL87k9bib7CzJ5csrfnXrNDfXT2Gu/UHt0N9cfUxvc3X4v8y37Gm/jyNktt/Sz70C4k08EekVFBdnZ2Tz44IOG97V69WpWr15teD/NwaHCXZRWHiOy07WEhXYlsHUwYP8T/UKQzHqpPxHhfd12ni7Yh0geGgN//rRp9zRv6Q+zRkG0gaehHT62n07tewIQ02sU/boP54U5H/HiW79k9oSlTi7NHryXruvOYb3JP7qX4f0n1Nk2mdkf1/Rd2+W25xe7NxDdbRjz73+HmS9G88uxzzXlZdfrp3vH/g2+9gu1GxHoYP+BCpsNvsi+ctuG3NzXvgx384lADw4OprraxDee9lElZUdY/OZ9daZ1COlM0oy3KTtVwoqNv2L+jHcACA5s1+AVkG88m8vbn75Aytd/42c3Gf+FfUG7IHh8HPxr5+X3xC7VuyNMvRE6hhhXG9S9rqGo5BBdOtiPvB4tzSe8bUTNvMttg0vVW9c2G2Cpt20aGmK80vYsPvEjUV2H4O/nT89OA4nrd0eTary0n0Zf+79rN4qfxf6XXNQ19oOklWcde16b1jD5ehgSaVhpl+UTgS7eqUNo5wYvFa+uruKFt2Yw566X6vwZfamScvu4ecd23TlZ4eTVVi7QugXcEwcjoux7Ypm5Dd9EyQL07WL/keiB3ewfdqN169iXrNwvAPjxyPdEdh5ItbUai6XuYa/GtsGlGlrXR0pySBhyX722tfsGx7ZneNtuHC8rxGq1knski6rqi/cUdrTGhvpp7LU3VrurDY2Evp3tPz24bT8cq2i4XXiw/Ueib4iyh7qnKNDF5dJ3rSc7/xtWvf8bAGaPX8q1PW+s127n/k94dMVNtPBvRdL0+ntr7hLRHu69wb5HVlxqH18/W2U/JfGaUPeeR3xB7y6xHD1pv5F2bvH3XBt5I+erznKy4ieOlxURFtrFqeU1tK7zj+4jquuQy/YNjm3Pm2Mm8vv1/48d+1O5PW4mf3v/aZ6dvtapGhvqp7HX3ljtRmjTGm691n6GVOlpyDt+8SyY0AD7hWVtA73jDhAWm07hcKumXPrvKlHXwMNjnX+eEZejf3cwjR37N/N/r/DjCO26QZzxO2Jeo/a6dtVpiw2t64OF3/HNvg+5b/RvGnyOq/o2wqW1N7f3yOVoD13ES01OeMqwZUd1jSWqa+O/nWdk31frSrU3Zwp08YjYqASP3ia2OdG6bj4U6G5m5A2AjOo7xKB7oHh7357Q3F6vK2idXaQxdBERk/DKS/9FRMR5CnQREZNQoIuImIQCXUTEJBToIiImoUAXETEJBbqIiEko0EVETEKBLiJiEgp0ERGTUKCLiJiEAl1ExCQU6CIiJqFAFxExCQW6iIhJKNANlp+fz2233caAAQMYNGgQzzzzjKdLEhGTUqAbrEWLFrz44ov88MMP7Nixg23btrFp0yZPlyUiJqSfoDNYly5d6NKlCwCtWrVi8ODB5OXlebgqETEj7aG70fHjx9m4cSNjx471dCkiYkIKdDc5e/YskyZN4rHHHqN///6eLkdETEg/Eu0G1dXV3HvvvfTo0YPf//73ni5HRExKge4Gs2fPxmq18tprr2GxWDxdjoiYlGmHXLKyskhMTCQuLo7WrVtjsVioqKhosG1SUhJ33HEH4eHhWCwWVqxY4bI6tm7dymuvvUZGRgZDhw5lyJAhvPrqqzXz9X0qIq5i2rNcMjMzSUlJYdiwYQQEBLB169ZG27766qsMGTKEu+66izVr1ri0jpEjRzYa2jabjdXJH9InMoJbhg92ab8i0vyYdg/9/vvvJz8/n40bNzJmzJjLti0tLWXLli0899xz7inu3w7+WMi+Q/m0aOHv1n5FxJx8OtB3797NPffcQ3h4OAEBAURHR5OUlASAn5/jL82Ztq5is9nYvDWT0OA2xA3u5/b+RcR8fHbIJTMzk1GjRhEZGcny5cvp0aMHOTk5bNu2zSP1/OeLf23ycxe8/JoLKxERM3nhN3Mdbuuzgf7kk08SEhLCV199RWhoaM302bNne7AqERHP8clAP3XqFFu2bOHhhx+uE+ae5My36IHcw6xa9z4/HzuSG68baGBVItKc+GSgnzhxAqvVSkREhKdLqdGUIZdNqVvZlNr42TciIs7sLPrkQdH27dvj5+fH4cOHPV2KiIjX8Mk99KCgIEaNGsXbb7/NwoULvWLYxZFvUZvNxl/W/pOSk+X8OvFeWrbwydUvIl7KZxNl+fLljBo1ihEjRvD0008TGRlJXl4eW7ZsYdWqVZw6dYqUlBQA9uzZA8DGjRsJCAigY8eOxMfH1ywrPT2do0ePcvToUQB27txJcnIyABMmTCAoKMglNR/8sZDcgiP8fOxIhbmIuJxP38tl165dLFiwgM8//5wzZ87QvXt37rvvPhYuXEhubi69evVq8Hnx8fGkpaXV/D8hIYH09PQG2+bk5NCzZ0+X1Juxax+fb/+Oh2dOVKCLiMv5dKD7IqvV6pELmUTE/BToIiImoV1FERGTUKCLiJiEAl1ExCQU6CIiJqFAFxExCQW6iIhJKNBFRExCgS4iYhIKdBERk1Cgi4iYhAJdRMQkFOgiIiahQBcRMQkFuoiISSjQRURMQoEuImISCnQREZNQoIuImIQCXUTEJBToIiImoUAXETEJBbqIiEko0EVETEKBLiJiEgp0ERGTUKCLiJiEAl1ExCT+F51Ck2YI0onAAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<Figure size 473.904x204.68 with 1 Axes>"
      ]
     },
     "execution_count": 18,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# U(φ) \n",
    "# Note that exp(aZ) = rZ(-a*2)\n",
    "# https://qiskit.org/documentation/locale/ja_JP/tutorials/circuits/3_summary_of_quantum_operations.html\n",
    "\n",
    "data = [alpha, beta]\n",
    "feature_map_zz_dim2([alpha,beta],rep=1).draw(output=\"mpl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 521.123x204.68 with 1 Axes>"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# U†(φ)\n",
    "data = [alpha, beta]\n",
    "feature_map_zz_dim2(data,rep=1).inverse().draw(output=\"mpl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 895.449x204.68 with 1 Axes>"
      ]
     },
     "execution_count": 19,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "#  U†(φ)U(φ)\n",
    "data = [alpha, beta]\n",
    "feature_map_zz_dim2([alpha,beta],rep=1).draw(output=\"mpl\")\n",
    "qc_u1 = feature_map_zz([alpha,beta],dim=2,rep=1)\n",
    "qc_u2 =  feature_map_zz([gamma, delta],dim=2,rep=1).inverse()\n",
    "qc = qc_u1.compose(qc_u2)\n",
    "\n",
    "qc.draw(output=\"mpl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "kernel is 0.1875\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "execution_count": 21,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# |<0|U†(φ)U(φ)>|^2\n",
    "qc_u1 = feature_map_zz([alpha,beta],dim=2,rep=1)\n",
    "qc_u2 =  feature_map_zz([gamma, delta],dim=2,rep=1).inverse()\n",
    "qc = qc_u1.compose(qc_u2).assign_parameters({alpha:1,beta:2,gamma:3,delta:4})\n",
    "qc.measure([0,1],[1,0])\n",
    "counts = execute(qc, backend , seed_simulator=seed).result().get_counts()\n",
    "kernel= counts[\"00\"]/sum(counts.values()) if \"00\" in counts else 0\n",
    "print(f\"kernel is {kernel}\")\n",
    "plot_histogram(counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {},
   "outputs": [],
   "source": [
    "# for dim >= 3\n",
    "def feature_map_zz(data, dim=2, rep=2):\n",
    "    q = QuantumRegister(dim)\n",
    "    c = ClassicalRegister(dim)\n",
    "    qc = QuantumCircuit(q,c)\n",
    "    for _ in range(rep):\n",
    "         # Π_i exp(i*a_i*Z_i) H_i\n",
    "        for i in range(dim):\n",
    "            qc.h(q[i])        \n",
    "            qc.rz(-2*data[i],q[i])\n",
    "         # Π_(i<j)  exp(i*(π-a_i)(π-a_j) Z_1*Z_2)\n",
    "        for i in range(dim):\n",
    "            for j in range(i+1, dim):\n",
    "                    qc.cx(i, j) \n",
    "                    qc.rz(-2*(np.pi-data[i])*(np.pi-data[j]),q[j])\n",
    "                    qc.cx(i,j)\n",
    "    return qc\n",
    "\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAuAAAADWCAYAAACQYjeQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlzAAALEgAACxIB0t1+/AAAADh0RVh0U29mdHdhcmUAbWF0cGxvdGxpYiB2ZXJzaW9uMy4xLjMsIGh0dHA6Ly9tYXRwbG90bGliLm9yZy+AADFEAAAgAElEQVR4nO3deVyVZf7/8dcBZVFERFAEFVABc0PFLEyFtBytRltMUyttVbOZpmzsN0N9p3LUTCdnpmZs0WYss0yzVa1cknLLrVRaxFzABXNBBVyBc35/3AOCbAc85z7nwPv5ePgQ73Od6/rc5764/Zz7vu7rsthsNhsiIiIiImIKL1cHICIiIiJSlygBFxERERExkRJwERERERETKQEXERERETGREnARERERERMpARcRERERMZEScBEREREREykBFxERERExkRJwERERERETKQEXERERETGREnARERERERMpARcRERERMZEScBEREREREykBFxERERExkRJwERERERETKQEXERERETGREnARERERERMpARcRERERMZEScBEREREREykBFxERERExkRJwERERERETKQEXERERETGREnARERERERMpARcRERERMZEScBEREREREykBFxERERExUT1XB1DXLNkCh066pu2IJnB7j5q9d9dqyD3q2Hjs0agZxPUzv10RERERZ1ECbrJDJ2GPCxLZK5V7FE4ddHUUIiIiIp5PQ1BEREREREykBFxERERExERKwEVERERETKQEXERERETERHoIUxxm4uxkfsrYgLd3fby8vAlrEs3I/ikkxd/p6tBERERE3IYScHGoUTc8w6gbnqawsICP17/CtAUjaRfRjYiQdq4OTURERMQtaAiKOIW3dz0GXfMQhdYC9hz+3tXhiIiIiLgNj07AV61axejRo4mJiaFBgwZERUVx7733sm/fvlLlVq5cSf/+/WnRogW+vr60aNGCW265hQ0bNrgo8tovv+Ain62fDUDLkFgXRyMiIiLiPjx6CMqrr77KqVOnmDRpErGxsRw4cIDJkyeTkJDAtm3biIqKAuDEiRPEx8czbtw4mjVrxpEjR5g1axZ9+/YlNTWVXr16uXZHapEFq6awKHUm5y7k4u1dnyfunEOb8C4ALN80l5Vb3y4um5W9l87RffjTyHdcFW6FbDbjb4vFtXGI+7LZPK9/eGLMYi71ERFzWGy2olTD8xw7dozQ0NBS2zIyMoiOjuaPf/wj06dPr/C9OTk5hIaGcu+99/LGG284O9RiL6+o3kqYi/+aTOtON9Dz1qft2l6Zts3gdzfa33ZJW96reiXMibOT6R5zA6NueJrcsyf526IH8K3vX26CnZ1zhCdfu54pDyyjRXB0hXUGtYQed9Us5uqy2uC7DFi7CzJOGNvahEKfOOjSSv8pibGSberPsD0TLhRAk4bQqx1cFwMNfF0dXfkuFsCGX2DdbjiWA/W8oWMEJLWH6NCq3y+1364s+HqX8XehDcKDoHcsXB1t9BcRcTy3HYJitVqZOXMmMTEx+Pn5ER8fT2pqKnFxcTz88MMAZZJvgMjISEJCQjh4sPJsMSAgAF9fX3x8fJwSf13XqEETnrhzDpt+Xsb6tI9LvWa1Wpn27ijuHzi10uTbTFYrvLMe3l4H+08YybjVZnxZ+s83sGjTpaviUjdtz4S/LYfNe43kG+DkGVi6HV76HE6ddW185Tl30fjS/+FWI/m2AfmFxr7840tYm+7qCMXVPt8Js1fDj4ehwGqc5w6fhIXfwutrjP4iIo7ntgn4/fffz+TJkxk7dizLly9n2LBhjBgxgr1795KQkFDh+9LS0jh27BidOnUq81phYSH5+flkZGQwYcIEAMaNG+e0fajrAhsEc0efJ3jz8z9jtVqLt7+94jmiwzrTu/NtLoyutNRdsHW/8XPJRLvox/W/wMY9Zkcl7uJEHry11vhSVt73sKLX3c3izXAg2/i5ZNy2Eq9nHDc7KnEXPx6Cz3cYP5d33ks/Akv1DL2IU7hlAr5gwQLmzZvHJ598wpNPPsn1119PSkoKiYmJFBQUVJiA5+fnM3bsWEJCQoqvkpeUlJSEj48PUVFRfPjhhyxfvpzOnTuXKvPrr78yYMAAGjRoQHx8PN99951T9rGuuK3PY2TnZLFi61sAbNu9iq3pX/LQzS+6OLJLrFZjWEFlLMCan3QVvK5av9u4NV8RG7D3mDFExV2cPmcMqaqMBfhGV8HrrDU/G32gMut/gQv5poQjUqe45UOY06ZNY+DAgSQlJZXa3q5dO+rXr18maQaw2Ww8+OCDbNq0iaVLl9K0adMyZebOncvp06c5dOgQc+bMYeDAgXz66ackJycXlxk/fjzt27fn448/5u2332bo0KGkp6fj7V31QDiLHYOE70j5ipZXJVdZrqRNH09h67KZpbbln8+jdacbqlVPauoafj/g+mq9p8jMcV8R3za50jJ/G7+mzLaGfoEsed64BJedc4RXPnqUqQ8sp349+4b+pKau4eoRNYvZXk3C23Pviz9VWsYG/JoDgSGtyMuuYjC81Dojp35PSKsuVf6OD7n3KbZ+5h5fLuN6jWTgI5U/4GwD1u7M5p7ryp4vpXazWLz43Vv5WCyVX4e7WACx3W8kM22lSZGJeK7qPFbpdgn4wYMHSUtL4/HHHy/zWmZmJh07dsTXt+zTTo8++ijz589nwYIFDBgwoNy64+Liin8eMmQIiYmJPPbYY2zfvh2A3Nxcli5dyqFDh/D39+fhhx9m6tSpbNy4keuuu85Be1h9PYeklPsQpqeZv3IyZ86fZsbCMcXbWoXG8Yehr7kuKMDbzi8D1S0rtYd3PV+7vmC7U/+wNxavevWdHIm4Iy/velUm38Vl3ahfi9QWbpmAA4SFhZXafu7cOVJTU7npppvKvGfixInMnj2buXPnMnz4cLva8fLyokePHrz55pvF23bv3k3Tpk0JCQkp3ta5c2d+/PFHuxJwe775VHcWFEdKSkpm8V9rNobCnllQqvL72//F72//V7Xek5SUjG22c8d9nLsIz3xgPIBUGd96cOzwHuprVoA65z/fwI4DVQ9Bev0fk4n/cLI5QVVh/3H4+xeVl7EAMRGNqnXVRmqP5z+C7DNVl9v89VJCGjk/HpG6xO3GgBclv+nppQcmvvjii2RlZdG9e/dS21NSUnjppZd4+eWXue++++xuJz8/n7Vr19Ku3aUl0s+cOUNgYGCpcoGBgeTl5VV3N8SD+PtAj+iqx0Je2w4l33XUdTGVJ98WIMAPOrU0LaQqRTY1ppOrrF/bMKabk7qpqmNvAdq3QMm3iBO43RXwNm3a0KVLF6ZOnUpwcDAREREsXryYZcuWAZR6AHPGjBlMnTqV4cOHk5CQwMaNG4tfCwwMpEOHDgDceuutdO3alfj4eIKDg8nMzOT1118nLS2NJUuWFL+nYcOG5ObmloonJyeHgIAAZ+6yuIFBXeCnw5BzruwsFxYguCHc2NEVkYk7iGluzIm8eV/Z14oS3LuuAW83uqRhscCdPeFfK6HQWv7sLbFh0D3K7MjEXVwXazyoWzRTTkkWwLc+3FrxpGMicgXcciGe9PR0xo4dy6ZNm2jatCmjR4+mUaNGpKSkkJOTg7+/PwDJycmkpqaWW0dSUhJr1qwBYPr06SxevJg9e/aQm5tLcHAwiYmJPPnkk/Tu3bv4Pbm5uYSEhHD48OHihzijo6OZP3++w8aAu3IIirMX4nEGMxfiOXnGmO/7x8OXtlks0LklDL0aAv3NiUPck9VqzJn89S44X2JWiOaN4dbucFW462KrzP7jsGQzZJZIsup5Q2JbGNxdd3XquvP5sGQLbNlnTLNZpF0zuONqaBHkuthEajO3TMDLc88997B9+3Z27Njh1HZuv/12WrduzQsvvMD8+fOZOnUqu3fvtmsWFHsoAa8eMxPwIifyYPL/1g569jYIamBu++LeLhbApIXGz3/4jTHUwxNWST10EmYYNxKZdqcx9EqkSN55ePoD4+c/3WJ8sRQR53G7ISgV2bJlC9dee63T25k9ezZ33303TZo0ISYmhg8++MBhyXdd8VPmt7z6yeN4eXkT27IH4wfPIu/cKfYcNlZ0aBvelQB/972s0rTEiCMl33I5nxJnzaiQisu5m4gml35W8i2XC/C79LOSbxHn84gEPC8vj/T0dB555BGnt9W8eXNWrFjh9HZqs+ZBkcwYuxqf+n5MWzCKfVk7+fVkBuvSPsSGjTPnc+jVcbCrwxQRERFxCY9IwAMCAigsLHR1GGKn4MBLU0h6e9XDy8u4g2Ar8RjYxYILTHtnJMdOH2BAjzHsPriVicPmmh6riIiIiNk8IgEXz7T38A5OnzlOZPMONA0Mx9/XGNvRNrwra3d+QEzLBJ6+533GTI/h3hufdW2wIiIiIiZRAi41lp1zhCnvlH5CMrhRGCl3v0fO2Wxe+ehRnr77fQAC/INKLWX/68kM2oZ3xdvLm6jmHekR9xszQxcRERFxGSXgUmPBgWH8bfyaMtsLCwt44d27eeiWGaWGo5QU0rglJ3IOY7Va2X8kjYLC/HLLiYiIiNQ2brRshNQWqTsWkX5gM3OWPsXE2cn8uH9DmTK9O9/Od7+sYuqCEQzoMYY3lk5yQaQiIiIi5tMVcHG4ft1G0K/biErL+Ps0JGXUuyZFJCIiIuI+dAVcRERERMREugJuspKLYXhS242aOS4OT2hXRERExFmUgJvs9h6ujqBm4vq5OgIRERGR2kFDUERERERETKQEXERERETERErARURERERMpARcRERERMRESsBFREREREykBFxERERExERKwEVERERETKQEXERERETERErARURERERMpARcRERERMRESsBFREREREykBFxERERExERKwEVERERETFTP1QHUNUu2wKGTrmk7ognc3qNm7921GnKPOjYeezRqBnH9zG/XVVz1OXuyutZHRETE8ykBN9mhk7DHAxOs3KNw6qCro6j99DmLiIjUfhqCIiIiIiJiIiXgIiIiIiIm0hAUkf85chp++RUOZl/atvBbaBkMMc2hWaDrYhPXO3sBfs6CAyX6x7y1EB4EkSHQrhl46ZKGiIjYQQm4OMzE2cn8lLEBb+/6eHl5E9YkmpH9U0iKv9PVoVXqh4Ow6kfYe6zsaxt+ufRzTHO4oSPEtTAvNlfw1OPoLCfy4Ms02LYf8gtLv/ZdhvEHoElD6BMLfeOgnrfpYYqIiAdRAi4ONeqGZxh1w9MUFhbw8fpXmLZgJO0iuhER0s7VoZVx5gJ8sBm2ZdhXfvevxp9r2sJtCeBX37nxuZInHUdnsdlg3W745Du4WFB1+ZNnjLKb98LIXtAq2PkxioiIZ9INU3EKb+96DLrmIQqtBew5/L2rwykj9xy8vML+5Lukb/fAKyuNBL62c/fj6Cw2G3y4FRZvti/5LinrNPzzS9iV5ZzYRETE8ykBF6fIL7jIZ+tnA9AyJNbF0ZSWXwivfmWM+a6pg9nwxhootDosLLfkzsfRmVakwde7av7+/EKYk1r6eQIREZEiHj0EZdWqVbz11lusX7+eQ4cO0axZM/r27ctzzz1HdHR0cbmVK1cybdo0fvzxR7KzswkODiYhIYGUlBQSExNduAe1z4JVU1iUOpNzF3Lx9q7PE3fOoU14FwCWb5rLyq1vF5fNyt5L5+g+/GnkO6bG+PmOqhdD+vso4+8/VBLa/uPG2PEBnRwXm7vwhOPoLAdOwOc7Ky9jT//IL4QFG+GJ32hMuIiIlObRCfirr77KqVOnmDRpErGxsRw4cIDJkyeTkJDAtm3biIqKAuDEiRPEx8czbtw4mjVrxpEjR5g1axZ9+/YlNTWVXr16uXZHapGR/VMYdcPT5J49yd8WPcD3v6xmUM8HABjU84Hin7NzjvDka9czZuBfTY3vWA6s/tFx9X2xE3q2gaAGjqvTHbj7cXSmxZvBanNMXYdPGuPIk9o7pj4RuST3vPFw9Omz4O8D3SIhpJGroxJ3UWiFtIOQeQK8LNC2OcSGGT+7A49OwP/9738TGhpaalufPn2Ijo5m9uzZTJ8+HYDhw4czfPjwUuUGDRpEaGgo//nPf9w6AV/812Rad7qBnrc+bdd2d9GoQROeuHMOo19oy/q0j+nVaUjxa1arlWnvjuL+gVNpERxdSS2Ot3Y3OCi3Aoxf8A2/wKAuDqzUjbjrcXSWzBOQccKxda5Nhz5x7nPSF/F0Vhss/R6++qn0l+Wl240kfMS14OPR2Y1cqfQj8PY640takRU/QEgA3NcXIpq4LrYibjsG3Gq1MnPmTGJiYvDz8yM+Pp7U1FTi4uJ4+OGHAcok3wCRkZGEhIRw8GDl63kHBATg6+uLj4+PU+IXCGwQzB19nuDNz/+M1XppsPTbK54jOqwzvTvfZmo8Nhts2ef4ejfvdXyd7sTdjqMzOaN/HMuFjOOOr1ekrvrsO2P4X3l3qr7LgP9+Y5zvpW7afxxe+wryzpd97cQZeGUFHM81P67LuW0Cfv/99zN58mTGjh3L8uXLGTZsGCNGjGDv3r0kJCRU+L60tDSOHTtGp05lB+YWFhaSn59PRkYGEyZMAGDcuHFO2weB2/o8RnZOFiu2vgXAtt2r2Jr+JQ/d/KLpsZzIc87MJdlnyv9Fr03c6Tg6k6OvfhfJdFK9InXNqbPw1c+Vl/nxMPxy1Jx4xP0s/R6s1vLvdttscD7f+ALnam55k2bBggXMmzePNWvWkJSUBMD111/Ptm3bWLJkSYUJeH5+PmPHjiUkJKT4KnlJSUlJrFu3DoDmzZuzfPlyOnfuXKrMX/7yFxYtWsTPP//M+++/z9ChQx28d7XX38avKbOtoV8gS543poLIzjnCKx89ytQHllO/nvl3Hg6fcm7dsWHOq99M7n4cnSmriodza8qZfU+kLtmyr+qr2xbg21+MxdOkbsnOM9brqIwN2LwPbu8B9V34gLxbJuDTpk1j4MCBxcl3kXbt2lG/fv0ySTOAzWbjwQcfZNOmTSxdupSmTZuWKTN37lxOnz7NoUOHmDNnDgMHDuTTTz8lOTm5uExMTAz/+Mc/eOaZZ6odt8VS9SDPO1K+ouVVyVWWK2nTx1PYumxmqW355/No3emGatWTmrqG3w+4vlrvKTJz3FfEt02u0XuLzF85mTPnTzNj4Zjiba1C4/jD0NcqfE9q6hquHlGzmC8Xd90oBo6fX2pb0WwWFano9ctnvxh08xD2bvvkCqIzOOJzdraaHEdnclgfsVh47O3S80o6qn+8/c5CRibedQXBOcZj843MxZ5zldQ9ntA/kke/Qud+Y/Hyrjh9sdpsfLQ8lXt6O+b/DvEcLWJ6Mewv66osV1AITUIjOHPysEPbt1Vj7JPbJeAHDx4kLS2Nxx9/vMxrmZmZdOzYEV9f3zKvPfroo8yfP58FCxYwYMCAcuuOi4sr/nnIkCEkJiby2GOPsX379uLtd999NwBTpky50l1xmJ5DUsp9CNPT/P72f/H72//lsvatBRedVndhQR1Yled/XH0cncZmo7AgH+96jl/itNCJfU+kLrl47jQWr8pHz9psVi6cvYKFHsRjXTxn33G32azkn3ftQHC3TMABwsJK388/d+4cqamp3HTTTWXeM3HiRGbPns3cuXPLzHZSES8vL3r06MGbb7555UH/jz3ffF5eAXtcNDYtKSmZxX+t2ZMpW96DU5U/1+oUSUnJ2GY75mmaQydhxrLS2yqax9meeZ5L2rL2c4dMf+Wqz9mTObKPvPBZ6QWaHNU/Jj5yD1/MvufKgnOAonirc5VG6g5P6B8Hs2Hm8srLeHl585ffDeHTl9x3P8Q5bDaY+qnxkGVFR99igavCvbhwNsfU2C7ndg9hhoSEAJCenl5q+4svvkhWVhbdu3cvtT0lJYWXXnqJl19+mfvuu8/udvLz81m7di3t2rW78qDFI4Q1ds6CKA18oGmA4+sV87UKdk69LZ1Ur0hd0zIY4sKMcd7lsVgguCF0bW1qWOImLBZjcbxKv3rZoF8HsyKqmNtdAW/Tpg1dunRh6tSpBAcHExERweLFi1m2zLh0WfIBzBkzZjB16lSGDx9OQkICGzduLH4tMDCQDh2MT/jWW2+la9euxMfHExwcTGZmJq+//jppaWksWbLE3B0Ul/H2gk4R8H2mY+vt0sr4pRfP16WV8XCOI/n7QDs9DCbiMKN7w2trjOk9LZbSD2UGNYDx/bX6bF12dRs4eQaW7TC+qJVMxi0WY554d3hA1+0ScC8vLxYtWsTYsWMZP348TZs2ZfTo0UyYMIGUlBS6dLm04snSpUsBWLhwIQsXLixVT1JSEmvWrAEgMTGRxYsX889//pPc3FyCg4NJTEwkNTWV3r17m7ZvNTH06TXV2u4Ofsr8llc/eRwvL29iW/Zg/OBZ5J07xZ7D3wPQNrwrDf0a8/xbQzmec4hWoXFMumueKbH1jnV8An5djGPrK7JozUy6xfSnXUS3K6rHZrOV+az3HN7O1vQvGZb8xzLHJsA/qFTb9hzPAP+gK97f8topT8nYHa1DhPEf+KmzjqvzmjZaFETEkRr4wu9vhB8Owaa9xmqHYCRW3SL1+yYwoDN0bmWsRLz2fwMq+neAXjHuc8faLbtpbGwsX331Valt99xzD1dddRX+/v7F24oS7Ko89dRTPPXUU3aVzc/Pp7CwEKvVSn5+PufPn8fX19etnwp3N82DIpkxdjU+9f2YtmAU+7J28uvJDNalfYgNG2fO5xDWJIoGfoG8PPoDUubeTHbOEYIDnT+PX9tm0CHcmCfWEbpHQquyE+5cMavVyg/713Fn8pNXXNe+rJ1lPuu24fG8/eWzWK1W0vatLXVsrr3qllJt23M8e3UcfMVxltdOdIuyMx6VjN2rioexqsvbC37bzVhBzREa+kL/jo6pS0Qu8fYy7lh1aXVp7Po1bV0bk7iXFkEw9OpLCfhvr+xalsO5ZQJeni1btnDttdc6vZ2HHnqIefOMq7HffPMNAPv27SMqKsrpbdcWJRNpb696eHkZ9wJtJW4E5Z47SUjjCC7mn+dU3lEaNTBnkKzFAsOvgReWwrkrnJiikR/ccbVj4rrc3qzthAS1BGD7njU8O+822rSI50j2PtpFdOO5MR/ZXVdFn3VESEzxVeySx6Zk22Df8bxYcIFp74zk2OkDDOgxht0HtzJx2Nxq7XN57VS070Wxx7TsXkmNNdM9ErZnwo4DV17X0KuNfiIiIlKSRyTgeXl5pKen88gjjzi9rf/+97/897//dXo7dcHewzs4feY4kc070DQwHH9f475P2/CuxYnf/TPaExESa+odhsYN4OFkeHU1XCgov0xVs1v414ex1xtXOJ3h0PHdNG8SBUDn6L7EterJCw99wfR37+WBm6ZVszYjUb78sw5r2oYDx36mZ/ubSh2brelfFrddUmXHc+3OD4hpmcDT97zPmOkx3HvjszXZ7TLttAptX+6+F8XujATcYoFRiZBzzljSuDz2zH5yc7xxO1xERORyHpGABwQEUFhY6Oow5DLZOUeY8k7pxUWCG4WRcvd75JzN5pWPHuXpu98HIMA/qNwFZub/eT/vrX6BZd++weBezv+CVSQ6FCbcAPPWGkvUV0fzQOMhoPAmzokNSk8DlpW9lxbBbQA4dvoAIY0jil+r7BhcrsxnbbMBljLHprwpyKo6nr+ezKBteFe8vbyJat6RHnG/qVGMl7dT4b7/L3Zn8a0P4/vBuxur/8xAfW+4tTtcF+uc2ERExPN5RAIu7ik4MKzcZcsLCwt44d27eeiWGZWO687OPQJAaFArTuWZPzl666Yw6Wb4fIfxoMbFCq6GF/GtB33jjIc7nL18bcvQWNL2rwUg48gPRIZ1pNBaiMVSesxzRcfgcuV91key95HctezqjCXbBvuOZ0jjlpzIOYzVamX/kTQKCvOrHWN57VS07xXF7ki+9WFMH/g+Az77Ho7b8UXtqnC4PQFCA50amoiIeDgl4OJwqTsWkX5gM3OWGg++PjBoGh2iEsuU+373Kh57pRf1vH1IGVX2aqgZfOvBkO7GvKFb98Mvv8KBbMg9b1xfDfQ35p2NaQ4JUUZSZoY2LeI5dsoYhLz/1x/oEJlIfsEFTuUd5UROFk0DW1SrvvI+6wPHdtE2vGulbYN9x7N359t5adGDbNu9ggE9xvDG0kn8edSCasVYXjsV7XtFsTtD10jo0hp2ZcHOA0b/OJYLhVbwq2/cCYlsCldHK/EWERH7KAEXh+vXbQT9uo2oulz3Udw38K8mRFQ1fx9jisLebjJswMvLi05Rvfnl0HeM7P/n4u1vTNxZo/ou/6z3HN5Oh6he5c4iUrLtdhHd7Dqe/j4NSRn1bo1iK46xnHZKJvpF+15Z7M7iZTGubl8VblqTIiJSiykBF3FTjpiCsCJtw+NpGx7vkravVFWxi4iIuDsl4OIS8W2Ty30oUxxPn7WIiIh7UQJusggnzpzhzLYbNXNcHJ7QrqvUtf11BH1mIiLiaZSAm+z2Hq6OoGbi+rk6grpBn7OIiEjtZ95TTCIiIiIiogRcRERERMRMSsBFREREREykBFxERERExERKwEVERERETKQEXERERETERErARURERERMpARcRERERMRESsBFREREREykBFxERERExERKwEVERERETKQEXERERETERErARURERERMVM/VAdQ1S7bAoZOuaTuiCdzeo2bv3bUaco86Nh57NGoGcf3Mb1dEHMdV5w9PpnOf+1O/rj7160uUgJvs0EnY44G/sLlH4dRBV0chIp5I5w+pjdSv5UpoCIqIiIiIiImUgIuIiIiImEgJuIhILVdQCAezL/375Bmw2VwXj7gXmw1O5F369+GTUGh1XTz2upAPGccv/fvMBdfFIu7HaoUjpy/9+1guWN3ovKcx4CIitdCFAti2HzbuMZLvkgnVcx9BgC90iIDesdC6qcvCFBex2Yznkdbthl1ZcPbipddeXAb1vCGyKSS2g66tjX+7g9zzsPEX2Loffj0NJfOplMUQ3BDiW8N1MRDSyFVRiqsUWmHnQdiwG/Ydg4uFl16b8gn41YeY5nBdLMSGgZfFdbEqAReHmTg7mZ8yNuDtXR8vL2/CmkQzsn8KSVdBk+sAABdvSURBVPF3ujo0kTrl+0xYvAnyKrkimHcBNu01/nSMgGE9oXED82I0k85NpR3PhXc3Vj4hQEGh8fqeo/DZ93DXtdC+hXkxXs5qhTU/w7LtUFDJ1fnsM/DVT7DmJ+gdB7d0Bd9amumoX5e296jRr4/lVlzmfL6RoO88aFx4GHEttAgyL8aSNARFHGrUDc/w6ZQ8ljx7gt9cPYZpC0Zy6Pgvrg5LpE4otBr/Af33m8qT78v9cAhe+Ax2H3FebK6mc5NheyZMX1q92bhOnYVXV8Mn21wzdOnsBXhlJXzyXeXJd0k24JtdMHOZ8YWjtlK/Nvrkl2nw8orKk+/LZZ6AmcuNixCuoARcnMLbux6DrnmIQmsBew5/7+pwRGo9qw3eWQ/f7qnZ+8/lw2tfwe5fHRuXu6nL56YdB4wvZ/mFVZctz+qf4MOt5ibh5/Nh9mrYe6xm7z+WayRmJce410Z1uV9/sdO4M1KTbllohQUban7evBIenYCvWrWK0aNHExMTQ4MGDYiKiuLee+9l3759pcqtXLmS/v3706JFC3x9fWnRogW33HILGzZscFHktV9+wUU+Wz8bgJYhsS6ORqT2W5sO2zIqL/P3UcafihRYYd5ayDvv2NjcSV09N53IM76gVZakVNU/AL7eZQxxMsuSLXAgu/IyVcV9+hy8tc4YxlJb1dV+/XMWfL6z8jL29Ov3N0HWKcfFZQ+PHhn16quvcurUKSZNmkRsbCwHDhxg8uTJJCQksG3bNqKiogA4ceIE8fHxjBs3jmbNmnHkyBFmzZpF3759SU1NpVevXq7dkVpkwaopLEqdybkLuXh71+eJO+fQJrwLAMs3zWXl1reLy2Zl76VzdB/+NPIdV4UrUiucyINPv3NMXXnnjauc91znmPrcRV0+N9lssPBb48FcR1i82XiQLcDPMfVV5KfDjhsekHEcUnfB9Vc5pj53UZf79YV8eG+jY+oquhL++EDzHsz06AT83//+N6GhoaW29enTh+joaGbPns306dMBGD58OMOHDy9VbtCgQYSGhvKf//zHrRPwxX9NpnWnG+h569N2bXe1kf1TGHXD0+SePcnfFj3A97+sZlDPBwAY1POB4p+zc47w5GvXM2bgX10ZrkitkPpzzYcVlGfbfhjUpXbNIlGXz02ZJyDdgeP7z1yADb/AjZ0cV2d5VqQ5tr6vfoQ+se4zo4sj1OV+vWWf8XyCoxzINmYEuirccXVWxm2HoFitVmbOnElMTAx+fn7Ex8eTmppKXFwcDz/8MECZ5BsgMjKSkJAQDh6sfH3YgIAAfH198fHxcUr8dV2jBk144s45bPp5GevTPi71mtVqZdq7o7h/4FRaBEe7KEKR2uFigeMfIrJhJFi1UV08N63b7fg61+927pzKWadqPu67IjnnjdkvaiP1a/etsyJum4Dff//9TJ48mbFjx7J8+XKGDRvGiBEj2Lt3LwkJCRW+Ly0tjWPHjtGpU9mv5oWFheTn55ORkcGECRMAGDdunNP2oa4LbBDMHX2e4M3P/4y1xOC7t1c8R3RYZ3p3vs2F0YnUDgeyjQfVHM2RV0zdTV07NznjWJ4869zZRZzV/9Sva4e883DYCWO2dx8xb7Eet0zAFyxYwLx58/jkk0948sknuf7660lJSSExMZGCgoIKE/D8/HzGjh1LSEhI8VXykpKSkvDx8SEqKooPP/yQ5cuX07lz5+LXL1y4wJgxY4iIiCAoKIh+/frx008/OW0/64Lb+jxGdk4WK7a+BcC23avYmv4lD938oosjE6kdqnpAraYOnzLmgq6t6sq5Kfe8Y2/Tl3TghHPqBef1a2fV6y7qSr921nG8UFC9qQyvhMVmc78FiTt37kzLli1Zvnx5qe1PPfUUs2bNIjc3F19f31Kv2Ww2xowZw4IFC1i6dCkDBgwoU++uXbs4ffo0hw4dYs6cOXz99dd8+umnJCcnA3DmzBlmzZrFfffdR1hYGNOnT+e9995jx44ddsVtsVQ9cv+OlK9oeVWyXfWBMdb7yJ5v8a5fen/zz+dx7e3PVmsM+MGf1vDBlOvtLl/SzHFfEd82uUbvLVI0Bm3qA8sJC46y6z3b96zhyVdrFrNIXdB7xIsk3PzHUtuqeuK/In+47NmsOY+Gc+ZUVg0ju8QR5w9nqsm5ydkcde4LjezKyCmln9B1VP/4ZsGTbFv2txpGVrnb/7SKVh37ldpWk7gvj/nMqSPMedQxKwqpX1efo/p1x6T7ueGhuaW2OapffzC1Hwd//KpGdVUnpXa7hzAPHjxIWloajz/+eJnXMjMz6dixY5nkG+DRRx9l/vz5LFiwoNzkGyAuLq745yFDhpCYmMhjjz3G9u3bAWjYsCFPP30pof3d735HSkoK58+fx8/PyY97V6LnkJRyH8L0NPNXTubM+dPMWDimeFur0Dj+MPQ11wUl4vGc+Mi+HRcVaoNafW5y4jG0WJx4E91Jcdtzoay2UL+uadXm9BG3uwK+ceNGEhMTWbp0KTfddFPx9nPnztG2bVtuuukm5syZU+o9EydOZNasWcydO5f77rvP7rYmTJjAm2++yblz58p9/YsvvmDcuHFl5hW/Ei+vqN4KZI6cBaVtM/jdjfa3XdKW9+CUCx5eCWoJPe4yv10RT7HyB2OpcHsUXSG6/IpPeSwWmD4MfBxwmcZV5w9P5qhz36mz8OyH9pWtTv8AYxnva9rWLK6q/Pcb++cbr07cYY3h/91S87hKUr+uPkf1650HYO7X9pWtbr+edBOEN6lZXNXhdmPAQ0JCAEhPTy+1/cUXXyQrK4vu3buX2p6SksJLL73Eyy+/XK3kOz8/n7Vr19KuXbtyXz958iQTJkxgypQp1dwDERHztAp2Tr1hgY5JvsW1GvtDw7I3jR3CWX0PoKWT6nZmzGIeZ/WPel7QvLFz6i7TljnN2K9NmzZ06dKFqVOnEhwcTEREBIsXL2bZsmUApR7AnDFjBlOnTmX48OEkJCSwceOlGdkDAwPp0KEDALfeeitdu3YlPj6e4OBgMjMzef3110lLS2PJkiVlYjh37hyDBw9m+PDhjBw50sl7LCJSc62aGv9pFDh4lb82zRxbn7iGxWLc/dxxwLH1NvBxbqLS1kn9T/26dghqAMENIfuMY+uNCgVvky5Nu10C7uXlxaJFixg7dizjx4+nadOmjB49mgkTJpCSkkKXLl2Kyy5duhSAhQsXsnDhwlL1JCUlsWbNGgASExNZvHgx//znP8nNzSU4OJjExERSU1Pp3bt3qfcVFBQwbNgwYmJi3OLq99Cn11Rru4jULQ18oFskbHbcSDkAEsu/OSgeKLGd4xPwa9s6N1GJCoEWjSHrtOPq9K0H3SMdV5+4jsVi9Oul2x1bby8Tz3tul4ADxMbG8tVXpZ9Aveeee7jqqqvw9/cv3laUYFflqaee4qmnnrKr7IMPPojVauX111+3O14p7afMb3n1k8fx8vImtmUPxg+eRd65U+w5bAxUbRvelQD/IBdHKVJ7JLU3VoVz1AM9sWHOu8W7aM1MusX0p11EN4fXvefwdramf8mw5D+We84paju/8KIp56jyzoVVxe0McS0gPMhx8ybX84brYh1TV0UsFri+g7E8uKNcFwu+9R1XX0me0q8b+jXm+beGcjznEK1C45h017xqx+Mu/TqxHaz60XHrIDQNgC6tHFOXPdxuDHhFtmzZUukCPI6QkZHBvHnzWL16NUFBQQQEBBAQEEBmpp1PgggAzYMimTF2NbMe+YZTeUfZl7WTtH1rWbn1bVZsfYsde+18ckJE7NIyGPp1cExd9b1hWE/H1HU5q9XKD/vXOSVJAWgbHs+P+9djtVrLnHNKtm3WOaq8dqqK2xm8LMYDk14OmtzhlngjWXG2q6MhLswxdYU0goGdqy5XE57Ur/dl7aSBXyAv/24jp88cJzun+isTuUu/DvCD2xyYFt51rfHl0ixueQX8cnl5eaSnp/PII484tZ3IyMhqzeEo5QsOvHTG9Paqh5eX0aNtJa7Ppcy9mZH9U+gY1YtP1v+bAP8g+nXTeHuRmhrYBfYdq3z5bntmARh+jZGsOMPerO2EBLUEjPmAn513G21axHMkex/tIrrx3JiP7K6ronNIREhM8RW/kueckm3bc466WHCBae+M5NjpAwzoMYbdB7cycVjpeYerUl47Fe13UdwxLbtXUmPNtWoKg7vDR1srLmNP/+gYAX3jqi7nCBYLjEiEv39R+WJCVcXtUw/uvc55DxV7Ur/OPXeSkMYRXMw/z6m8ozRqUP1bXe7Ur3u2MVav3LK/4jL29OsBnSCmucPCsotHJOABAQEUFtbiJdlqqb2Hd3D6zHEim3egaWA4/r7GJZO24V2JbdmDN5f/mY5RvVj/w8f85d4PXBytiGer7w0PXw9vrKneVKdFvCww7BroEe3w0IodOr6b5k2iAOgc3Ze4Vj154aEvmP7uvTxw07Rq1fX40DfKPYeENW3DgWM/07P9TaXOOVvTvyxuu0hl56i1Oz8gpmUCT9/zPmOmx3Dvjc/WeL9LttMqtH25+10Ut7MSFYDk9lBYCJ/aOW3l5Tq1hNG9wcvEe+dBDeDRG2D2ajiRV/33+9WHh5OhdVOHh1bMk/p1URJ//4z2RITEXtGc1+7Qr4u+pGExhuHVxA0dYVCXqss5mkck4OKesnOOMOWd0hN6BjcKI+Xu98g5m80rHz3K03e/D0CAf1CpFcMC/IPIOXucrOx9NGoQXHyCEJGa86sPj/Q3xkV+sRMK7bzzG9YYRiY6N0mB0qvEZWXvpUVwGwCOnT5ASOOI4tcqO7cUCWkcXv45xGYDLGXOOZff3azqHPXryQzahnfF28ubqOYd6RH3m1LvtyfG8tqpcL//F7ez9e9oHOf3vrU/oa3vDYO7GWOoHTWMpTpCGsGTg+CjbfDtHvvf176FcUenSUPnxQae1a+LzP/zft5b/QLLvn2Dwb0ujS7wxH7t7QWjEo0r2B9utX9MeGN/46JDx4iqyzqDEnCpseDAMP42fk2Z7YWFBbzw7t08dMuMUreqLndVZCL/+uh33NTzISdGKVK3eHsZt1O7R8K63UbCcvZi+WUjQ6BPLHRtbc7Yx5ahsaTtXwtAxpEfiAzrSKG1sMyKihWdWy5X3jnkSPY+kruWXemjZNv2nKNCGrfkRM5hrFYr+4+kUVBY+n91e2Isr52K9ruiuJ0hJgyeutm4Yrh2Nxw+WX65QD+4th30ijGuRLuSv48xjv26GPgmHb7LgIJyboxbLNAhHHrHGgm4GYsaelK/BsjONcZ9hwa14lRe6dtlntqvLRZjUagOEbDxF1i/G05WMGypeaDxZbJnG+OihasoAReHS92xiPQDm5mz1Jh55oFB0+gQlVimXK8Og1mcOpP/0/ATEYcLaQRDusNvu8GxHDh00kjEvSzGA3Qtg523QEtF2rSI59gpYz68/b/+QIfIRPILLnAq7ygncrJoGtiiWvWVdw45cGwXbcO7Vtq2Peeo3p1v56VFD7Jt9woG9BjDG0sn8edRC6oVX3ntVLTfFcXtLD71jMS6VwzknocDJ4xx1jab0S9aBhv9xN1Wbm/d1LjaObynMUVh1im4WGB8gWweaKxg6GtyZuNJ/Rrg+92reOyVXtTz9iFl1HtlXq+KO/frRn5wYydjWEn2GTiYDXnnjX7c2B9aNjX+dgdKwMXh+nUbQb9uI6osF92iM5HNO+JTz+QsQKQO8bIYC6aYtbpbpbF4edEpqje/HPqOkf3/XLz9jYnlz6JQlcvPIXsOb6dDVC+8yhmkXLJte85R/j4NSRn1bo3iKlJeOyUToqL9rixuMzTyM64cepJ63saqlu6wsqUn9WuAft1Hcd/Av9YoNvCMfm3534UGM2brqSkl4CIiYpo7k590Wt1tw+NpGx7vkravRFVxi/tTvy5L/bpySsDFpf4+Ya2rQxARD6ZziNRGzurX8W2Ty30oU8ynBNxkEU08s+1GzRwXhye0KyKOo9/j6tNn5v50jKpPn9klFptWnhERERERMY3HLEUvIiIiIlIbKAEXERERETGREnARERERERMpARcRERERMZEScBEREREREykBFxERERExkRJwERERERETKQEXERERETGREnARERERERMpARcRERERMZEScBEREREREykBFxERERExkRJwERERERETKQEXERERETFRPVcHUFfceOONHD16FJvNRmxsLG+++SaBgYGuDktERERETGax2Ww2VwdRF5w+fZrGjRsD8MQTTxAYGMizzz7r2qBERERExHQagmKSouTbarVy5swZLBaLiyMSEREREVdQAm6iwYMHExYWxq5du5g4caKrwxERERERF9AQFJNZrVZSUlIICQlREi4iIiJSBykBd4Fdu3YxdOhQdu7c6epQRERERMRkGoJigpycHLKysor//cEHH9CxY0cXRiQiIiIirlKrE/C0tDTGjh1Ljx498PX1xWKxkJeXV27ZjIwM7rrrLoKDg2nQoAG9evVi9erVDonj9OnTDB48mM6dO9OlSxe+//57/vGPfzikbhERERHxLLV6HvCtW7eybNkyEhIS8PPzY926deWWy87Opnfv3tSvX5+XX36ZoKAgXnvtNQYOHMiqVavo06fPFcXRqlUrNm/eXO5rNptNM6KIiIiI1CG1egy41WrFy8u4yP/ss8/y3HPPkZubS0BAQKlyU6ZM4ZlnnmHnzp3FQ0MKCwvp0qULgYGBbNiwwWkxrtuSxp7MQ9z12/741K/V34dEREREhFowBGXnzp3ccccdhISE4OfnR0xMDCkpKQDFyXdVNm7cSGRkZKlx2d7e3gwcOJCNGzdy+PBhp8Sen1/Amo3fcf7CRSXfIiIiInWER2d9W7dupW/fvkRGRjJz5kxat27Nvn37WL9+fbXqyc/Px8fHp8x2X19fwBhLHh4e7pCYS/p2+0/knjnHXYP7O7xuEREREXFPHp2AT5w4kUaNGrFx40YCAwOLtz/wwAPVqqd9+/asXr2aX3/9lebNmxdvL0rkT5w4YVc9/2/669Vqt8gb735Wo/eJiIiIiHt44amH7S7rsUNQzp49yzfffMNdd91VKvmuiQcffBCAu+++m/T0dI4ePcrzzz/P2rVrAfuHsoiIiIiIVMVjr4CfPHkSq9VKRETEFdfVqVMnFi1axPjx44mLiwMgLi6O559/npSUFLuHn9j7zSc/v4AXX3uX0KZBPDzitzWOW0REREQ8j8cm4E2aNMHLy4tDhw45pL4hQ4bw29/+lvT0dLy8vIiJiWH69On4+/vTvXt3u+qo7hCU3DPnajxsRURERETcR50YgtKgQQP69u3Le++9R05OjkPq9PLyon379sTGxpKXl8cbb7zBmDFjaNiwoUPqFxERERHx6HnAS86CMmnSJCIjI8nMzOSbb75hzpw5nD17lmXLlgHw/vvvs2jRIt5++238/PwIDQ0lKSkJMOYLf/zxx0lOTiYoKIj09HRmzJiBn58fa9euJSgoyGExr92yk89WbeChEbfQtrXjZ1YREREREffm0Qk4wI4dO3jmmWf4+uuvOX/+PK1ateKuu+7i+eefZ//+/URHR5f7vqSkJNasWQMYCfiQIUPYvHkz2dnZtGjRgjvuuIP/+7//c2jyrbHfIiIiIuLxCbgnOXriFG8v+YJbf9NHV79FRERE6igl4CazWq2a1lBERESkDlMCLiIiIiJiIl2KFRERERExkRJwERERERETKQEXERERETGREnARERERERMpARcRERERMZEScBEREREREykBFxERERExkRJwERERERETKQEXERERETGREnARERERERMpARcRERERMZEScBEREREREykBFxERERExkRJwERERERETKQEXERERETGREnARERERERMpARcRERERMZEScBEREREREykBFxERERExkRJwERERERETKQEXERERETGREnARERERERMpARcRERERMZEScBEREREREykBFxERERExkRJwERERERETKQEXERERETHR/wdv2NKNOlnQ2AAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<Figure size 955.94x264.88 with 1 Axes>"
      ]
     },
     "execution_count": 25,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "data = [alpha, beta, gamma]\n",
    "feature_map_zz(data, dim=3,rep=1).draw(output=\"mpl\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {},
   "outputs": [],
   "source": [
    "def quantum_kernel(data_1, data_2, rep=2):\n",
    "    if len(data_1) != len(data_2):\n",
    "        print(\"Input size is insconsist\")\n",
    "        raise Exception\n",
    "    dim = len(data_1)\n",
    "    qc = feature_map_zz(data_1,dim=dim,rep=rep).compose(feature_map_zz(data_2,dim=dim,rep=rep).inverse())\n",
    "    qc.measure([0,1],[1,0])\n",
    "    qc.draw(output=\"mpl\")\n",
    "    counts = execute(qc, backend , seed_simulator=seed).result().get_counts()\n",
    "    kernel= counts[\"00\"]/sum(counts.values()) if \"00\" in counts else 0\n",
    "    return kernel"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "kernel is 0.0419921875\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 504x360 with 1 Axes>"
      ]
     },
     "execution_count": 27,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Quantum Kernel \n",
    "qc = feature_map_zz([3,4],dim=2,rep=2).compose(feature_map_zz([1,2],dim=2,rep=2).inverse())\n",
    "qc.measure([0,1],[1,0])\n",
    "qc.draw(output=\"mpl\")\n",
    "counts = execute(qc, backend , seed_simulator=seed).result().get_counts()\n",
    "kernel= counts[\"00\"]/sum(counts.values()) if \"00\" in counts else 0\n",
    "print(f\"kernel is {kernel}\")\n",
    "plot_histogram(counts)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Generate Sample data\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 28,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 5.29 s, sys: 14.4 ms, total: 5.3 s\n",
      "Wall time: 5.31 s\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "training_dataset_size=20\n",
    "testing_dataset_size=10\n",
    "feature_dim=2\n",
    "aqua_globals.random_seed = seed\n",
    "sample_Total, training_input_unnormalized, test_input_unnormalized, class_labels = ad_hoc_data(\n",
    "    training_size=training_dataset_size, \n",
    "    test_size=testing_dataset_size, \n",
    "    n=feature_dim, gap=0.3, plot_data=True\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 29,
   "metadata": {},
   "outputs": [],
   "source": [
    "records = []\n",
    "for label, data in training_input_unnormalized.items():\n",
    "    for x, y in data:\n",
    "        records.append({\"x\":x, \"y\":y, \"label\":label})\n",
    "df_train = pd.DataFrame(records)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>3.769911</td>\n",
       "      <td>0.628319</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>4.398230</td>\n",
       "      <td>5.780530</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.890265</td>\n",
       "      <td>0.314159</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.398230</td>\n",
       "      <td>3.895575</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>2.953097</td>\n",
       "      <td>0.565487</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>3.581416</td>\n",
       "      <td>6.094690</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>2.953097</td>\n",
       "      <td>0.314159</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>0.251327</td>\n",
       "      <td>6.220353</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>0.376991</td>\n",
       "      <td>2.827433</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>1.570796</td>\n",
       "      <td>1.884956</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>3.958407</td>\n",
       "      <td>5.340708</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>2.010619</td>\n",
       "      <td>5.089380</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.942478</td>\n",
       "      <td>0.502655</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.633628</td>\n",
       "      <td>1.884956</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>1.445133</td>\n",
       "      <td>1.507964</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>2.324779</td>\n",
       "      <td>2.261947</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>4.461062</td>\n",
       "      <td>1.068142</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.753982</td>\n",
       "      <td>6.220353</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>3.455752</td>\n",
       "      <td>0.816814</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>5.089380</td>\n",
       "      <td>1.130973</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>20</th>\n",
       "      <td>2.073451</td>\n",
       "      <td>2.953097</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>21</th>\n",
       "      <td>4.712389</td>\n",
       "      <td>5.969026</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>22</th>\n",
       "      <td>4.649557</td>\n",
       "      <td>0.314159</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>23</th>\n",
       "      <td>3.141593</td>\n",
       "      <td>2.324779</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>24</th>\n",
       "      <td>4.963716</td>\n",
       "      <td>2.701770</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>25</th>\n",
       "      <td>0.251327</td>\n",
       "      <td>4.649557</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>26</th>\n",
       "      <td>3.832743</td>\n",
       "      <td>4.649557</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>27</th>\n",
       "      <td>0.062832</td>\n",
       "      <td>0.753982</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>28</th>\n",
       "      <td>4.900885</td>\n",
       "      <td>0.691150</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>29</th>\n",
       "      <td>4.272566</td>\n",
       "      <td>1.507964</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>30</th>\n",
       "      <td>2.701770</td>\n",
       "      <td>4.084070</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>31</th>\n",
       "      <td>0.816814</td>\n",
       "      <td>1.382301</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>32</th>\n",
       "      <td>3.141593</td>\n",
       "      <td>1.005310</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>33</th>\n",
       "      <td>5.026548</td>\n",
       "      <td>5.780530</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>34</th>\n",
       "      <td>5.152212</td>\n",
       "      <td>2.890265</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>35</th>\n",
       "      <td>2.261947</td>\n",
       "      <td>2.953097</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>36</th>\n",
       "      <td>4.209734</td>\n",
       "      <td>5.906194</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>37</th>\n",
       "      <td>3.581416</td>\n",
       "      <td>1.570796</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>38</th>\n",
       "      <td>1.068142</td>\n",
       "      <td>2.513274</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>39</th>\n",
       "      <td>2.199115</td>\n",
       "      <td>2.890265</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           x         y label\n",
       "0   3.769911  0.628319     A\n",
       "1   4.398230  5.780530     A\n",
       "2   2.890265  0.314159     A\n",
       "3   4.398230  3.895575     A\n",
       "4   2.953097  0.565487     A\n",
       "5   3.581416  6.094690     A\n",
       "6   2.953097  0.314159     A\n",
       "7   0.251327  6.220353     A\n",
       "8   0.376991  2.827433     A\n",
       "9   1.570796  1.884956     A\n",
       "10  3.958407  5.340708     A\n",
       "11  2.010619  5.089380     A\n",
       "12  0.942478  0.502655     A\n",
       "13  1.633628  1.884956     A\n",
       "14  1.445133  1.507964     A\n",
       "15  2.324779  2.261947     A\n",
       "16  4.461062  1.068142     A\n",
       "17  0.753982  6.220353     A\n",
       "18  3.455752  0.816814     A\n",
       "19  5.089380  1.130973     A\n",
       "20  2.073451  2.953097     B\n",
       "21  4.712389  5.969026     B\n",
       "22  4.649557  0.314159     B\n",
       "23  3.141593  2.324779     B\n",
       "24  4.963716  2.701770     B\n",
       "25  0.251327  4.649557     B\n",
       "26  3.832743  4.649557     B\n",
       "27  0.062832  0.753982     B\n",
       "28  4.900885  0.691150     B\n",
       "29  4.272566  1.507964     B\n",
       "30  2.701770  4.084070     B\n",
       "31  0.816814  1.382301     B\n",
       "32  3.141593  1.005310     B\n",
       "33  5.026548  5.780530     B\n",
       "34  5.152212  2.890265     B\n",
       "35  2.261947  2.953097     B\n",
       "36  4.209734  5.906194     B\n",
       "37  3.581416  1.570796     B\n",
       "38  1.068142  2.513274     B\n",
       "39  2.199115  2.890265     B"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_train"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 31,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "bae7271e2ceb4b5a9e53b625f51f4198",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=40.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "q_kernel = []\n",
    "for ind, row in tqdm(df_train.iterrows(),total=len(df_train)):\n",
    "    q_kernel.append([quantum_kernel(row[:2].values, row_in[:2].values) for ind_in, row_in in df_train.iterrows()])\n",
    "q_kernel =  np.array(q_kernel)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 33,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fdfc656dad0>"
      ]
     },
     "execution_count": 33,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x648 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12, 9))\n",
    "sns.heatmap(np.array(q_kernel), annot=False, fmt='g', cmap='Blues')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 32,
   "metadata": {},
   "outputs": [],
   "source": [
    "# Visualize Kernel calculated from Library"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 34,
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 7.98 s, sys: 245 ms, total: 8.23 s\n",
      "Wall time: 6.38 s\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fdf43678310>"
      ]
     },
     "execution_count": 34,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x648 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "os.environ['QISKIT_IN_PARALLEL'] = 'TRUE'  # brokeProcessErrorはこれで治るっぽい\n",
    "shots=1024\n",
    "# Wall time: 9.47 s for train_size = 20 test_size = 10\n",
    "backend = BasicAer.get_backend('qasm_simulator')\n",
    "feature_map = ZZFeatureMap(feature_dim, reps=2)\n",
    "svm = QSVM(feature_map, training_input_unnormalized, test_input_unnormalized, None)  # the data for prediction can be fed later.\n",
    "svm.random_seed = seed\n",
    "quantum_instance = QuantumInstance(backend, shots=shots, seed_simulator=seed, seed_transpiler=seed)\n",
    "result = svm.run(quantum_instance)\n",
    "kernel_matrix = result['kernel_matrix_training']\n",
    "\n",
    "plt.figure(figsize=(12, 9))\n",
    "sns.heatmap(np.array(kernel_matrix), annot=False, fmt='g', cmap='Blues')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 35,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "diff of kernel value between scratch and library, 0.008418718985919948\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fdf42c30d50>"
      ]
     },
     "execution_count": 35,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x648 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "print(f\"Difference of kernel value between scratch and library, {abs(kernel_matrix - q_kernel).max()}\")\n",
    "plt.figure(figsize=(12, 9))\n",
    "sns.heatmap(abs(kernel_matrix - q_kernel), annot=False, fmt='g', cmap='Blues')"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Optimization after Getting Kernel Matrix"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 36,
   "metadata": {},
   "outputs": [],
   "source": [
    "# polynominal kernel\n",
    "power=2\n",
    "def polynomial_kernel(x, y):\n",
    "    return (1 + np.dot(x, y)) ** power\n",
    "\n",
    "poly_kernel = [] \n",
    "for i in range(len(df_train)):\n",
    "    poly_kernel.append(\n",
    "        [polynomial_kernel(df_train.iloc[i,:2], df_train.iloc[ii,:2]) for ii in range(len(df_train))]\n",
    "    )\n",
    "poly_kernel = np.array(poly_kernel)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 37,
   "metadata": {
    "code_folding": []
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 984 ms, sys: 4.07 ms, total: 988 ms\n",
      "Wall time: 989 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "# Gaussian Kernel\n",
    "sigma = 0.3\n",
    "# def gaussian_kernel(x, y, sigma):\n",
    "def gaussian_kernel(x, y):\n",
    "    return np.exp(-norm(x-y)**2 / (2 * (sigma ** 2)))\n",
    "\n",
    "\n",
    "g_kernel = [] \n",
    "for i in range(len(df_train)):\n",
    "    g_kernel.append(\n",
    "#         [gaussian_kernel(df_train.iloc[i,:2], df_train.iloc[ii,:2], sigma=0.5) for ii in range(len(df_train))]\n",
    "            [gaussian_kernel(df_train.iloc[i,:2], df_train.iloc[ii,:2]) for ii in range(len(df_train))]\n",
    "    )\n",
    "g_kernel = np.array(g_kernel)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 38,
   "metadata": {},
   "outputs": [],
   "source": [
    "# prediction function\n",
    "def f(x, a, t, X, b, kernel_func):\n",
    "#     print(x)\n",
    "    sum_value = 0.0\n",
    "    for n in range(len(X)):\n",
    "        sum_value += a[n] * t[n] * kernel_func(x, X[n])\n",
    "    return sum_value + b\n",
    "   \n",
    "def get_laglange_param(t, kernel,print_mode=True):\n",
    "    # Calculate Lagrange parameter \"a\" by Quadratic Programming\n",
    "    N = len(t)\n",
    "    K = np.zeros((N, N))\n",
    "\n",
    "    for i in range(N):\n",
    "        for j in range(N):\n",
    "            K[i, j] = t[i] * t[j] * kernel[i, j]\n",
    "    \n",
    "    P = cvxopt.matrix(K)\n",
    "    q = cvxopt.matrix(-np.ones(N))\n",
    "    temp1 = np.diag([-1.0]*N)\n",
    "    temp2 = np.identity(N)\n",
    "    G = cvxopt.matrix(np.vstack((temp1, temp2)))\n",
    "    temp1 = np.zeros(N)\n",
    "    temp2 = np.ones(N) * C\n",
    "    h = cvxopt.matrix(np.hstack((temp1, temp2)))\n",
    "    A = cvxopt.matrix(t, (1,N))\n",
    "    b = cvxopt.matrix(0.0)\n",
    "    sol = cvxopt.solvers.qp(P=P, q=q, G=G, h=h, A=A, b=b)\n",
    "    a = array(sol['x']).reshape(N)\n",
    "    if print_mode:\n",
    "        print(\"Lagrange Param\")\n",
    "        print(a)\n",
    "    \n",
    "    # extract support vector index\n",
    "    S = []\n",
    "    M = []\n",
    "    for n in range(len(a)):\n",
    "        if 0 < a[n]:\n",
    "            S.append(n)\n",
    "        if 0 < a[n] < C:\n",
    "            M.append(n)\n",
    "    \n",
    "    # Caclulate intercept b\n",
    "    sum_value = 0\n",
    "    for n in M:\n",
    "        temp = 0\n",
    "        for m in S:\n",
    "            temp += a[m] * t[m] * kernel[n, m]\n",
    "        sum_value += (t[n] - temp)\n",
    "    b = sum_value / len(M)\n",
    "#     print(\"Support Vector and Intercept\")\n",
    "#     print(S, b)\n",
    "    return a, b\n",
    "\n",
    "def draw_contour(X, t, kernel_func, mesh=30):\n",
    "    # draw training data \n",
    "    for n in range(len(X)):\n",
    "        if t[n] > 0:\n",
    "            scatter(X[n,0], X[n,1], c='b', marker='o')\n",
    "        else:\n",
    "            scatter(X[n,0], X[n,1], c='r', marker='o')\n",
    "\n",
    "    # draw support vector\n",
    "    #    for n in S:\n",
    "    #        scatter(X[n,0], X[n,1], s=80, c='c', marker='o')\n",
    "    \n",
    "    grid_min = X.min() - abs(X.min())*0.1\n",
    "    grid_max = X.max() + abs(X.max())*0.1\n",
    "    # draw boundary\n",
    "    X1, X2 = meshgrid(linspace(grid_min, grid_max,mesh), linspace(grid_min,grid_max,mesh))\n",
    "    w, h = X1.shape\n",
    "    X1.resize(X1.size)\n",
    "    X2.resize(X2.size)\n",
    "    Z = array([f(array([x1, x2]), a, t, X, b, kernel_func) for (x1, x2) in tqdm(zip(X1, X2),total=len(X1))])\n",
    "    X1.resize((w, h))\n",
    "    X2.resize((w, h))\n",
    "    Z.resize((w, h))\n",
    "    CS = contour(X1, X2, Z, [0.0], colors='k', linewidths=1, origin='lower')\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 39,
   "metadata": {},
   "outputs": [],
   "source": [
    "kernel_dict = {\"gaussian\":{\"mat\":g_kernel, \"func\":gaussian_kernel},\n",
    "                         \"polynominal\": {\"mat\":poly_kernel, \"func\":polynomial_kernel},\n",
    "                         \"quantum\": {\"mat\": q_kernel, \"func\": quantum_kernel}\n",
    "                        }"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gaussian Ver"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 43,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -1.4034e+01 -3.9285e+01  2e+02  4e+00  3e-16\n",
      " 1: -1.1054e+01 -3.0303e+01  2e+01  5e-16  4e-16\n",
      " 2: -1.2515e+01 -1.4156e+01  2e+00  1e-15  4e-16\n",
      " 3: -1.3526e+01 -1.3616e+01  9e-02  2e-16  3e-16\n",
      " 4: -1.3584e+01 -1.3591e+01  7e-03  2e-15  2e-16\n",
      " 5: -1.3588e+01 -1.3588e+01  4e-04  6e-16  2e-16\n",
      " 6: -1.3588e+01 -1.3588e+01  2e-05  4e-16  2e-16\n",
      " 7: -1.3588e+01 -1.3588e+01  4e-07  1e-15  2e-16\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[0.5        0.5        0.4999999  0.5        0.5        0.5\n",
      " 0.20369032 0.5        0.5        0.36572421 0.5        0.5\n",
      " 0.5        0.49999896 0.5        0.5        0.5        0.5\n",
      " 0.5        0.5        0.48615509 0.5        0.5        0.5\n",
      " 0.49999999 0.5        0.5        0.5        0.5        0.5\n",
      " 0.5        0.5        0.5        0.5        0.49999999 0.49996545\n",
      " 0.5        0.5        0.5        0.08329286]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "915e6846241c45a5854da47a4c3780c5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "mesh=10\n",
    "C = 0.5        # hyper-parameter of soft-margin SVM\n",
    "\n",
    "\n",
    "# load traing dataset as input\n",
    "N = len(df_train)\n",
    "data = df_train.iloc[:,0:2].values\n",
    "label = df_train.iloc[:,2].apply(lambda x: 1.0 if x==\"A\" else -1.0)\n",
    " \n",
    "# decide the choice of kernel\n",
    "kernel_choice = kernel_dict[\"gaussian\"]\n",
    "\n",
    "a, b = get_laglange_param(t=label, kernel=kernel_choice[\"mat\"])\n",
    "draw_contour(X=data,t=label, kernel_func=kernel_choice[\"func\"],mesh=mesh)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 42,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -1.4034e+01 -3.9285e+01  2e+02  4e+00  3e-16\n",
      " 1: -1.1054e+01 -3.0303e+01  2e+01  5e-16  4e-16\n",
      " 2: -1.2515e+01 -1.4156e+01  2e+00  1e-15  4e-16\n",
      " 3: -1.3526e+01 -1.3616e+01  9e-02  2e-16  3e-16\n",
      " 4: -1.3584e+01 -1.3591e+01  7e-03  2e-15  2e-16\n",
      " 5: -1.3588e+01 -1.3588e+01  4e-04  6e-16  2e-16\n",
      " 6: -1.3588e+01 -1.3588e+01  2e-05  4e-16  2e-16\n",
      " 7: -1.3588e+01 -1.3588e+01  4e-07  1e-15  2e-16\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[0.5        0.5        0.4999999  0.5        0.5        0.5\n",
      " 0.20369032 0.5        0.5        0.36572421 0.5        0.5\n",
      " 0.5        0.49999896 0.5        0.5        0.5        0.5\n",
      " 0.5        0.5        0.48615509 0.5        0.5        0.5\n",
      " 0.49999999 0.5        0.5        0.5        0.5        0.5\n",
      " 0.5        0.5        0.5        0.5        0.49999999 0.49996545\n",
      " 0.5        0.5        0.5        0.08329286]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "ec42effc032349d0a2e639412597c7f5",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=900.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "mesh=30\n",
    "C = 0.5        # hyper-parameter of soft-margin SVM\n",
    "\n",
    "\n",
    "# load traing dataset as input\n",
    "N = len(df_train)\n",
    "data = df_train.iloc[:,0:2].values\n",
    "label = df_train.iloc[:,2].apply(lambda x: 1.0 if x==\"A\" else -1.0)\n",
    " \n",
    "# decide the choice of kernel\n",
    "kernel_choice = kernel_dict[\"gaussian\"]\n",
    "\n",
    "a, b = get_laglange_param(t=label, kernel=kernel_choice[\"mat\"])\n",
    "draw_contour(X=data,t=label, kernel_func=kernel_choice[\"func\"],mesh=mesh)\n"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Quantum Ver"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "C = 0.5        # hyper-parameter of soft-margin SVM\n",
    "mesh = 10\n",
    "\n",
    "# load traing dataset as input\n",
    "N = len(df_train)\n",
    "data = df_train.iloc[:,0:2].values\n",
    "label = df_train.iloc[:,2].apply(lambda x: 1.0 if x==\"A\" else -1.0)\n",
    " \n",
    "# decide the choice of kernel\n",
    "kernel_choice = kernel_dict[\"quantum\"]\n",
    "\n",
    "a, b = get_laglange_param(t=label, kernel=kernel_choice[\"mat\"])\n",
    "draw_contour(X=data,t=label, kernel_func=kernel_choice[\"func\"],mesh=mesh)\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 179,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -1.6261e+01 -4.2721e+01  2e+02  4e+00  9e-16\n",
      " 1: -9.8006e+00 -3.3896e+01  2e+01  6e-16  1e-15\n",
      " 2: -1.1119e+01 -1.3550e+01  2e+00  6e-16  7e-16\n",
      " 3: -1.1991e+01 -1.2431e+01  4e-01  8e-16  7e-16\n",
      " 4: -1.2189e+01 -1.2259e+01  7e-02  4e-16  9e-16\n",
      " 5: -1.2226e+01 -1.2231e+01  5e-03  7e-16  8e-16\n",
      " 6: -1.2229e+01 -1.2229e+01  2e-04  7e-16  8e-16\n",
      " 7: -1.2229e+01 -1.2229e+01  6e-06  2e-16  9e-16\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[4.99999717e-01 3.28829011e-01 4.99999973e-01 1.71166050e-01\n",
      " 4.99999963e-01 4.99999990e-01 4.99999969e-01 5.18658435e-05\n",
      " 1.55087165e-07 1.76241338e-04 4.99999981e-01 1.11577986e-06\n",
      " 4.99999985e-01 4.99999986e-01 4.99999965e-01 4.99999973e-01\n",
      " 4.99999966e-01 4.99999991e-01 4.99999961e-01 4.99999980e-01\n",
      " 4.99999960e-01 4.18337876e-01 4.99999959e-01 2.19603358e-06\n",
      " 4.99999749e-01 4.99999987e-01 3.47917999e-06 8.18803814e-02\n",
      " 8.54638675e-07 4.99999789e-01 4.99999950e-01 4.99999903e-01\n",
      " 4.99999918e-01 4.99999871e-01 4.99999920e-01 4.99999982e-01\n",
      " 3.24362352e-07 4.99999888e-01 4.99999864e-01 4.99999987e-01]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "960b5dd2cbc4442ba2375c4ad4656c18",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=900.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1h 24min 51s, sys: 20min 36s, total: 1h 45min 28s\n",
      "Wall time: 2h 30min 7s\n"
     ]
    }
   ],
   "source": [
    "C = 0.5        # hyper-parameter of soft-margin SVM\n",
    "mesh = 10\n",
    "\n",
    "# load traing dataset as input\n",
    "N = len(df_train)\n",
    "data = df_train.iloc[:,0:2].values\n",
    "label = df_train.iloc[:,2].apply(lambda x: 1.0 if x==\"A\" else -1.0)\n",
    " \n",
    "# decide the choice of kernel\n",
    "kernel_choice = kernel_dict[\"quantum\"]\n",
    "\n",
    "a, b = get_laglange_param(t=label, kernel=kernel_choice[\"mat\"])\n",
    "draw_contour(X=data,t=label, kernel_func=kernel_choice[\"func\"],mesh=mesh)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 30,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fa4b43edd10>"
      ]
     },
     "execution_count": 30,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "# 正解(Ground Truth)の領域\n",
    "plt.figure(figsize=(9, 6))\n",
    "sns.heatmap(np.array(sample_Total).T, annot=False, fmt='g', cmap='Blues').invert_yaxis()\n",
    "# # plt.scatter(df_train[])\n",
    "df_train[\"x_norm\"] = df_train[\"x\"].apply(lambda x: x/2/np.pi*100)\n",
    "df_train[\"y_norm\"] = df_train[\"y\"].apply(lambda x: x/2/np.pi*100)\n",
    "sns.scatterplot(data=df_train, x=\"x_norm\", y=\"y_norm\",hue=\"label\",s=70)\n",
    "# plt.show()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Evaluate Model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>label</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.565487</td>\n",
       "      <td>5.152212</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.513274</td>\n",
       "      <td>0.188496</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.439823</td>\n",
       "      <td>0.188496</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.021239</td>\n",
       "      <td>0.439823</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3.518584</td>\n",
       "      <td>0.816814</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.691150</td>\n",
       "      <td>0.691150</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.319469</td>\n",
       "      <td>3.330088</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.884956</td>\n",
       "      <td>3.455752</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>3.769911</td>\n",
       "      <td>0.565487</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>3.330088</td>\n",
       "      <td>6.031858</td>\n",
       "      <td>A</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>5.089380</td>\n",
       "      <td>3.330088</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.628319</td>\n",
       "      <td>1.319469</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.188496</td>\n",
       "      <td>3.958407</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.256637</td>\n",
       "      <td>0.565487</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>3.958407</td>\n",
       "      <td>4.586725</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>4.272566</td>\n",
       "      <td>2.827433</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.319469</td>\n",
       "      <td>6.094690</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.314159</td>\n",
       "      <td>3.958407</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>4.021239</td>\n",
       "      <td>1.319469</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>6.094690</td>\n",
       "      <td>4.084070</td>\n",
       "      <td>B</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           x         y label\n",
       "0   0.565487  5.152212     A\n",
       "1   2.513274  0.188496     A\n",
       "2   0.439823  0.188496     A\n",
       "3   4.021239  0.439823     A\n",
       "4   3.518584  0.816814     A\n",
       "5   0.691150  0.691150     A\n",
       "6   1.319469  3.330088     A\n",
       "7   1.884956  3.455752     A\n",
       "8   3.769911  0.565487     A\n",
       "9   3.330088  6.031858     A\n",
       "10  5.089380  3.330088     B\n",
       "11  0.628319  1.319469     B\n",
       "12  0.188496  3.958407     B\n",
       "13  1.256637  0.565487     B\n",
       "14  3.958407  4.586725     B\n",
       "15  4.272566  2.827433     B\n",
       "16  1.319469  6.094690     B\n",
       "17  0.314159  3.958407     B\n",
       "18  4.021239  1.319469     B\n",
       "19  6.094690  4.084070     B"
      ]
     },
     "execution_count": 48,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "records = []\n",
    "for label, data in test_input_unnormalized.items():\n",
    "    for x, y in data:\n",
    "        records.append({\"x\":x, \"y\":y, \"label\":label})\n",
    "df_test = pd.DataFrame(records)\n",
    "df_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -1.4034e+01 -3.9285e+01  2e+02  4e+00  3e-16\n",
      " 1: -1.1054e+01 -3.0303e+01  2e+01  5e-16  4e-16\n",
      " 2: -1.2515e+01 -1.4156e+01  2e+00  1e-15  4e-16\n",
      " 3: -1.3526e+01 -1.3616e+01  9e-02  2e-16  3e-16\n",
      " 4: -1.3584e+01 -1.3591e+01  7e-03  2e-15  2e-16\n",
      " 5: -1.3588e+01 -1.3588e+01  4e-04  6e-16  2e-16\n",
      " 6: -1.3588e+01 -1.3588e+01  2e-05  4e-16  2e-16\n",
      " 7: -1.3588e+01 -1.3588e+01  4e-07  1e-15  2e-16\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[0.5        0.5        0.4999999  0.5        0.5        0.5\n",
      " 0.20369032 0.5        0.5        0.36572421 0.5        0.5\n",
      " 0.5        0.49999896 0.5        0.5        0.5        0.5\n",
      " 0.5        0.5        0.48615509 0.5        0.5        0.5\n",
      " 0.49999999 0.5        0.5        0.5        0.5        0.5\n",
      " 0.5        0.5        0.5        0.5        0.49999999 0.49996545\n",
      " 0.5        0.5        0.5        0.08329286]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "0ce8a26a26c149d8afdd9a9e292d5f08",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=20.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# gaussian\n",
    "kernel_choice = kernel_dict[\"gaussian\"]\n",
    "data = df_train.iloc[:,:2].values\n",
    "label = df_train.iloc[:,2].apply(lambda x: 1.0 if x==\"A\" else -1.0)\n",
    "\n",
    "a, b = get_laglange_param(t=label, kernel=kernel_choice[\"mat\"])\n",
    "test_pred_g = array([f(array(row), a, label, data, b, kernel_choice[\"func\"]) for row in tqdm(df_test.iloc[:,:2].values,total=len(df_test))])\n",
    "\n",
    "\n",
    "df_test[\"pred_g\"] = test_pred_g\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -2.4271e+01 -4.7494e+01  3e+02  5e+00  8e-13\n",
      " 1: -1.2021e+01 -3.8705e+01  3e+01  9e-02  7e-13\n",
      " 2: -1.3330e+01 -1.7094e+01  4e+00  1e-02  5e-13\n",
      " 3: -1.4641e+01 -1.5334e+01  7e-01  2e-03  6e-13\n",
      " 4: -1.4850e+01 -1.4956e+01  1e-01  2e-04  7e-13\n",
      " 5: -1.4886e+01 -1.4912e+01  3e-02  4e-05  6e-13\n",
      " 6: -1.4896e+01 -1.4902e+01  6e-03  8e-06  8e-13\n",
      " 7: -1.4899e+01 -1.4899e+01  4e-04  1e-15  8e-13\n",
      " 8: -1.4899e+01 -1.4899e+01  1e-05  4e-16  9e-13\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[4.99999767e-01 4.99999913e-01 2.19786004e-07 4.99999963e-01\n",
      " 4.98928778e-01 4.99996573e-01 2.29605336e-07 2.55749867e-06\n",
      " 4.99999957e-01 4.99999925e-01 4.99999923e-01 4.99999874e-01\n",
      " 4.99998794e-01 4.99999924e-01 4.99999906e-01 4.99999933e-01\n",
      " 4.99999935e-01 2.54798744e-07 4.99999802e-01 4.99999956e-01\n",
      " 4.99999188e-01 4.99999841e-01 4.99999916e-01 4.99999155e-01\n",
      " 6.81638170e-08 3.60101624e-01 2.88855305e-01 4.99999876e-01\n",
      " 4.99999720e-01 3.13837405e-01 4.99999561e-01 4.99999843e-01\n",
      " 4.99999920e-01 9.13502434e-02 5.73799349e-08 4.99999161e-01\n",
      " 4.99999905e-01 4.99999773e-01 4.44786377e-01 4.99999247e-01]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "7f22419825164c86b4be2792d83cc926",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=20.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# polynominal\n",
    "kernel_choice = kernel_dict[\"polynominal\"]\n",
    "data = df_train.iloc[:,:2].values\n",
    "label = df_train.iloc[:,2].apply(lambda x: 1.0 if x==\"A\" else -1.0)\n",
    "\n",
    "a, b = get_laglange_param(t=label, kernel=kernel_choice[\"mat\"])\n",
    "test_pred_p = array([f(array(row), a, label, data, b, kernel_choice[\"func\"]) for row in tqdm(df_test.iloc[:,:2].values,total=len(df_test))])\n",
    "\n",
    "\n",
    "df_test[\"pred_p\"] = test_pred_p\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -6.6195e+00 -3.2162e+01  1e+02  3e+00  3e-16\n",
      " 1: -5.8688e+00 -2.3288e+01  2e+01  4e-16  4e-16\n",
      " 2: -6.3244e+00 -8.1191e+00  2e+00  1e-16  4e-16\n",
      " 3: -6.8881e+00 -7.2596e+00  4e-01  8e-16  5e-16\n",
      " 4: -7.0238e+00 -7.0923e+00  7e-02  2e-16  4e-16\n",
      " 5: -7.0518e+00 -7.0607e+00  9e-03  5e-16  4e-16\n",
      " 6: -7.0555e+00 -7.0559e+00  4e-04  4e-16  4e-16\n",
      " 7: -7.0556e+00 -7.0557e+00  1e-05  2e-15  4e-16\n",
      " 8: -7.0556e+00 -7.0556e+00  2e-07  7e-16  4e-16\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[2.51038711e-09 4.99999999e-01 4.99999994e-01 4.64709226e-01\n",
      " 4.99999999e-01 5.77246300e-09 1.60822009e-08 1.59506362e-09\n",
      " 1.89399157e-09 1.09471141e-01 4.99999991e-01 4.99999965e-01\n",
      " 4.99999997e-01 4.99999994e-01 4.54102318e-09 4.99999995e-01\n",
      " 4.99999987e-01 4.18223969e-01 7.80088970e-09 4.99999986e-01\n",
      " 1.18850931e-08 1.08716690e-09 4.99999997e-01 4.99999998e-01\n",
      " 4.94392308e-09 3.50711845e-01 4.26061145e-01 4.99999998e-01\n",
      " 4.99999994e-01 4.99999999e-01 4.99999904e-01 4.99999982e-01\n",
      " 3.50508821e-01 2.35003767e-01 7.26245036e-09 4.39143108e-01\n",
      " 4.22756193e-02 4.90599204e-01 1.58100859e-01 1.55820901e-08]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1203aa67087f487babea81e2d9ac54ba",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=20.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "# quantum\n",
    "kernel_choice = kernel_dict[\"quantum\"]\n",
    "data = df_train.iloc[:,:2].values\n",
    "label = df_train.iloc[:,2].apply(lambda x: 1.0 if x==\"A\" else -1.0)\n",
    "\n",
    "a, b = get_laglange_param(t=label, kernel=kernel_choice[\"mat\"])\n",
    "test_pred_q = array([f(array(row), a, label, data, b, kernel_choice[\"func\"]) for row in tqdm(df_test.iloc[:,:2].values,total=len(df_test))])\n",
    "\n",
    "\n",
    "df_test[\"pred_q\"] = test_pred_q\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 53,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>label</th>\n",
       "      <th>pred_g</th>\n",
       "      <th>pred_p</th>\n",
       "      <th>pred_q</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.565487</td>\n",
       "      <td>5.152212</td>\n",
       "      <td>A</td>\n",
       "      <td>-0.096395</td>\n",
       "      <td>0.091029</td>\n",
       "      <td>0.030786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.513274</td>\n",
       "      <td>0.188496</td>\n",
       "      <td>A</td>\n",
       "      <td>0.321576</td>\n",
       "      <td>2.078905</td>\n",
       "      <td>0.873018</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.439823</td>\n",
       "      <td>0.188496</td>\n",
       "      <td>A</td>\n",
       "      <td>0.005856</td>\n",
       "      <td>1.399000</td>\n",
       "      <td>1.403786</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>4.021239</td>\n",
       "      <td>0.439823</td>\n",
       "      <td>A</td>\n",
       "      <td>0.263573</td>\n",
       "      <td>0.990231</td>\n",
       "      <td>1.053747</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>3.518584</td>\n",
       "      <td>0.816814</td>\n",
       "      <td>A</td>\n",
       "      <td>0.627001</td>\n",
       "      <td>0.756279</td>\n",
       "      <td>1.224935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.691150</td>\n",
       "      <td>0.691150</td>\n",
       "      <td>A</td>\n",
       "      <td>0.176046</td>\n",
       "      <td>0.798123</td>\n",
       "      <td>0.652803</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>1.319469</td>\n",
       "      <td>3.330088</td>\n",
       "      <td>A</td>\n",
       "      <td>-0.045794</td>\n",
       "      <td>-0.649387</td>\n",
       "      <td>0.456991</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>1.884956</td>\n",
       "      <td>3.455752</td>\n",
       "      <td>A</td>\n",
       "      <td>-0.190002</td>\n",
       "      <td>-0.515379</td>\n",
       "      <td>0.230357</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>3.769911</td>\n",
       "      <td>0.565487</td>\n",
       "      <td>A</td>\n",
       "      <td>0.668612</td>\n",
       "      <td>0.971552</td>\n",
       "      <td>1.404014</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>3.330088</td>\n",
       "      <td>6.031858</td>\n",
       "      <td>A</td>\n",
       "      <td>0.316055</td>\n",
       "      <td>1.344614</td>\n",
       "      <td>1.012408</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>5.089380</td>\n",
       "      <td>3.330088</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.238775</td>\n",
       "      <td>-2.434111</td>\n",
       "      <td>-0.767631</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>0.628319</td>\n",
       "      <td>1.319469</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.424618</td>\n",
       "      <td>-0.017735</td>\n",
       "      <td>-1.113979</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>0.188496</td>\n",
       "      <td>3.958407</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.060730</td>\n",
       "      <td>-1.127349</td>\n",
       "      <td>-0.946742</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>1.256637</td>\n",
       "      <td>0.565487</td>\n",
       "      <td>B</td>\n",
       "      <td>0.254692</td>\n",
       "      <td>1.283304</td>\n",
       "      <td>-0.455718</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>3.958407</td>\n",
       "      <td>4.586725</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.441423</td>\n",
       "      <td>-0.773341</td>\n",
       "      <td>-0.968189</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>4.272566</td>\n",
       "      <td>2.827433</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.064848</td>\n",
       "      <td>-1.469338</td>\n",
       "      <td>-0.927049</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>1.319469</td>\n",
       "      <td>6.094690</td>\n",
       "      <td>B</td>\n",
       "      <td>0.051803</td>\n",
       "      <td>1.744376</td>\n",
       "      <td>-0.827565</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>0.314159</td>\n",
       "      <td>3.958407</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.060666</td>\n",
       "      <td>-1.029154</td>\n",
       "      <td>-0.839124</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>4.021239</td>\n",
       "      <td>1.319469</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.274184</td>\n",
       "      <td>-0.213170</td>\n",
       "      <td>-0.818061</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>6.094690</td>\n",
       "      <td>4.084070</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.026641</td>\n",
       "      <td>-3.787319</td>\n",
       "      <td>-1.003486</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           x         y label    pred_g    pred_p    pred_q\n",
       "0   0.565487  5.152212     A -0.096395  0.091029  0.030786\n",
       "1   2.513274  0.188496     A  0.321576  2.078905  0.873018\n",
       "2   0.439823  0.188496     A  0.005856  1.399000  1.403786\n",
       "3   4.021239  0.439823     A  0.263573  0.990231  1.053747\n",
       "4   3.518584  0.816814     A  0.627001  0.756279  1.224935\n",
       "5   0.691150  0.691150     A  0.176046  0.798123  0.652803\n",
       "6   1.319469  3.330088     A -0.045794 -0.649387  0.456991\n",
       "7   1.884956  3.455752     A -0.190002 -0.515379  0.230357\n",
       "8   3.769911  0.565487     A  0.668612  0.971552  1.404014\n",
       "9   3.330088  6.031858     A  0.316055  1.344614  1.012408\n",
       "10  5.089380  3.330088     B -0.238775 -2.434111 -0.767631\n",
       "11  0.628319  1.319469     B -0.424618 -0.017735 -1.113979\n",
       "12  0.188496  3.958407     B -0.060730 -1.127349 -0.946742\n",
       "13  1.256637  0.565487     B  0.254692  1.283304 -0.455718\n",
       "14  3.958407  4.586725     B -0.441423 -0.773341 -0.968189\n",
       "15  4.272566  2.827433     B -0.064848 -1.469338 -0.927049\n",
       "16  1.319469  6.094690     B  0.051803  1.744376 -0.827565\n",
       "17  0.314159  3.958407     B -0.060666 -1.029154 -0.839124\n",
       "18  4.021239  1.319469     B -0.274184 -0.213170 -0.818061\n",
       "19  6.094690  4.084070     B -0.026641 -3.787319 -1.003486"
      ]
     },
     "execution_count": 53,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "df_test"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Breast Cancer Dataset"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "CPU times: user 1.1 s, sys: 25 ms, total: 1.12 s\n",
      "Wall time: 154 ms\n"
     ]
    }
   ],
   "source": [
    "%%time\n",
    "\n",
    "training_dataset_size=20\n",
    "testing_dataset_size=10\n",
    "feature_dim=2\n",
    "aqua_globals.random_seed = seed\n",
    "sample_Total, training_input_unnormalized_br, test_input_unnormalized_br, class_labels = breast_cancer(\n",
    "    training_size=training_dataset_size, \n",
    "    test_size=testing_dataset_size, \n",
    "    n=feature_dim, plot_data=True\n",
    ")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "2dbe32f3f20c408cbd449e20c59f06f3",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=40.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "records = []\n",
    "for label, data in training_input_unnormalized_br.items():\n",
    "    for x, y in data:\n",
    "        records.append({\"x\":x, \"y\":y, \"label\":label})\n",
    "df_train_br = pd.DataFrame(records)\n",
    "\n",
    "q_kernel_br = []\n",
    "for ind, row in tqdm(df_train_br.iterrows(),total=len(df_train)):\n",
    "    q_kernel_br.append([quantum_kernel(row[:2].values, row_in[:2].values) for ind_in, row_in in df_train_br.iterrows()])\n",
    "q_kernel_br =  np.array(q_kernel_br)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {},
   "outputs": [],
   "source": [
    "poly_kernel_br = [] \n",
    "for i in range(len(df_train_br)):\n",
    "    poly_kernel_br.append(\n",
    "        [polynomial_kernel(df_train_br.iloc[i,:2], df_train_br.iloc[ii,:2]) for ii in range(len(df_train_br))]\n",
    "    )\n",
    "poly_kernel_br = np.array(poly_kernel_br)  \n",
    "\n",
    "g_kernel_br = [] \n",
    "for i in range(len(df_train_br)):\n",
    "    g_kernel_br.append(\n",
    "#         [gaussian_kernel(df_train_br.iloc[i,:2], df_train_br.iloc[ii,:2], sigma=0.5) for ii in range(len(df_train_br))]\n",
    "            [gaussian_kernel(df_train_br.iloc[i,:2], df_train_br.iloc[ii,:2]) for ii in range(len(df_train_br))]\n",
    "    )\n",
    "g_kernel_br = np.array(g_kernel_br)  "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 51,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -1.6261e+01 -4.2721e+01  2e+02  4e+00  9e-16\n",
      " 1: -9.8006e+00 -3.3896e+01  2e+01  6e-16  1e-15\n",
      " 2: -1.1119e+01 -1.3550e+01  2e+00  6e-16  7e-16\n",
      " 3: -1.1991e+01 -1.2431e+01  4e-01  8e-16  7e-16\n",
      " 4: -1.2189e+01 -1.2259e+01  7e-02  4e-16  9e-16\n",
      " 5: -1.2226e+01 -1.2231e+01  5e-03  7e-16  8e-16\n",
      " 6: -1.2229e+01 -1.2229e+01  2e-04  7e-16  8e-16\n",
      " 7: -1.2229e+01 -1.2229e+01  6e-06  2e-16  9e-16\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[4.99999717e-01 3.28829011e-01 4.99999973e-01 1.71166050e-01\n",
      " 4.99999963e-01 4.99999990e-01 4.99999969e-01 5.18658435e-05\n",
      " 1.55087165e-07 1.76241338e-04 4.99999981e-01 1.11577986e-06\n",
      " 4.99999985e-01 4.99999986e-01 4.99999965e-01 4.99999973e-01\n",
      " 4.99999966e-01 4.99999991e-01 4.99999961e-01 4.99999980e-01\n",
      " 4.99999960e-01 4.18337876e-01 4.99999959e-01 2.19603358e-06\n",
      " 4.99999749e-01 4.99999987e-01 3.47917999e-06 8.18803814e-02\n",
      " 8.54638675e-07 4.99999789e-01 4.99999950e-01 4.99999903e-01\n",
      " 4.99999918e-01 4.99999871e-01 4.99999920e-01 4.99999982e-01\n",
      " 3.24362352e-07 4.99999888e-01 4.99999864e-01 4.99999987e-01]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "21521c4864604f68be8ed6d5bff8a7a7",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=20.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -7.0511e+00 -4.1967e+01  3e+02  5e+00  5e-16\n",
      " 1: -4.3307e+00 -3.2479e+01  3e+01  6e-16  6e-16\n",
      " 2: -4.9028e+00 -8.1638e+00  3e+00  3e-16  5e-16\n",
      " 3: -5.5998e+00 -6.3486e+00  7e-01  1e-15  4e-16\n",
      " 4: -5.7942e+00 -5.9394e+00  1e-01  1e-15  4e-16\n",
      " 5: -5.8489e+00 -5.8613e+00  1e-02  4e-16  3e-16\n",
      " 6: -5.8530e+00 -5.8535e+00  5e-04  2e-16  4e-16\n",
      " 7: -5.8532e+00 -5.8532e+00  1e-05  2e-16  5e-16\n",
      " 8: -5.8532e+00 -5.8532e+00  1e-07  4e-16  5e-16\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[1.10903341e-09 9.06616526e-09 5.00000000e-01 1.13414736e-08\n",
      " 3.53249342e-01 4.99999999e-01 4.99999999e-01 7.32615124e-06\n",
      " 1.06093210e-01 1.47564321e-02 3.50596007e-01 1.44167620e-01\n",
      " 4.25891867e-01 3.57300253e-01 2.00866662e-09 4.99999999e-01\n",
      " 5.00000000e-01 2.37856149e-02 2.16130892e-09 2.49507247e-09\n",
      " 1.29920042e-01 1.75962967e-09 4.99999931e-01 1.28038122e-09\n",
      " 4.99999999e-01 4.99999999e-01 1.21171964e-09 1.26909940e-09\n",
      " 1.49045809e-09 1.86493887e-09 3.23320548e-01 3.38730614e-09\n",
      " 4.99999996e-01 2.87689102e-09 3.22607169e-01 5.00000000e-01\n",
      " 1.36984067e-09 3.25653223e-09 4.99999995e-01 4.99999999e-01]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "952b1467834b484792305c89fa6d3ac6",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=20.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -9.3807e+00 -3.7617e+01  2e+02  4e+00  1e-15\n",
      " 1: -5.4570e+00 -2.8362e+01  3e+01  2e-01  2e-15\n",
      " 2: -5.7332e+00 -8.8163e+00  3e+00  2e-02  1e-15\n",
      " 3: -6.7925e+00 -7.4308e+00  7e-01  3e-03  1e-15\n",
      " 4: -6.9988e+00 -7.2117e+00  2e-01  7e-04  2e-15\n",
      " 5: -7.0892e+00 -7.1126e+00  2e-02  3e-05  1e-15\n",
      " 6: -7.0995e+00 -7.1004e+00  9e-04  7e-07  2e-15\n",
      " 7: -7.1000e+00 -7.1000e+00  3e-05  1e-08  2e-15\n",
      " 8: -7.1000e+00 -7.1000e+00  5e-07  1e-10  2e-15\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[1.96332992e-09 1.11585177e-09 4.99999998e-01 1.21922224e-09\n",
      " 4.99999990e-01 4.99999997e-01 4.99999996e-01 5.12880152e-10\n",
      " 1.29742817e-09 5.12231101e-10 4.99999976e-01 4.92146066e-10\n",
      " 3.75330055e-10 4.31746198e-10 1.05052439e-09 4.99999998e-01\n",
      " 4.99999998e-01 4.99960609e-01 4.99999983e-01 4.99999552e-01\n",
      " 4.45391137e-09 2.45175035e-09 4.99999993e-01 4.00409220e-09\n",
      " 4.99999996e-01 4.99999997e-01 3.79021667e-09 3.73027543e-09\n",
      " 5.28275695e-09 4.99999962e-01 9.76127884e-10 4.99960182e-01\n",
      " 4.99999997e-01 2.60728669e-09 4.99999994e-01 4.99999999e-01\n",
      " 2.36779942e-09 1.94520551e-09 4.99999956e-01 4.99999998e-01]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1903dff61a7e4871a7e970c67305c167",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=20.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>label</th>\n",
       "      <th>pred_q</th>\n",
       "      <th>pred_g</th>\n",
       "      <th>pred_p</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.103898</td>\n",
       "      <td>0.179248</td>\n",
       "      <td>A</td>\n",
       "      <td>1.667739</td>\n",
       "      <td>0.819567</td>\n",
       "      <td>1.309670</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.530864</td>\n",
       "      <td>-0.457307</td>\n",
       "      <td>A</td>\n",
       "      <td>-0.155607</td>\n",
       "      <td>0.073109</td>\n",
       "      <td>-0.222972</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.210400</td>\n",
       "      <td>-0.424927</td>\n",
       "      <td>A</td>\n",
       "      <td>0.926252</td>\n",
       "      <td>1.369297</td>\n",
       "      <td>1.066187</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.021150</td>\n",
       "      <td>-0.468933</td>\n",
       "      <td>A</td>\n",
       "      <td>0.791680</td>\n",
       "      <td>1.234316</td>\n",
       "      <td>1.707608</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.208184</td>\n",
       "      <td>-0.487275</td>\n",
       "      <td>A</td>\n",
       "      <td>1.099681</td>\n",
       "      <td>1.354076</td>\n",
       "      <td>1.088123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.157078</td>\n",
       "      <td>-0.476244</td>\n",
       "      <td>A</td>\n",
       "      <td>0.049546</td>\n",
       "      <td>1.044874</td>\n",
       "      <td>2.214529</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.116275</td>\n",
       "      <td>0.720881</td>\n",
       "      <td>A</td>\n",
       "      <td>0.342349</td>\n",
       "      <td>0.542134</td>\n",
       "      <td>1.735123</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>-0.068440</td>\n",
       "      <td>-0.523605</td>\n",
       "      <td>A</td>\n",
       "      <td>1.166646</td>\n",
       "      <td>1.272920</td>\n",
       "      <td>1.579304</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-0.544330</td>\n",
       "      <td>-0.513921</td>\n",
       "      <td>A</td>\n",
       "      <td>-0.489345</td>\n",
       "      <td>0.028090</td>\n",
       "      <td>-0.287809</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>-0.310342</td>\n",
       "      <td>-0.539510</td>\n",
       "      <td>A</td>\n",
       "      <td>0.671727</td>\n",
       "      <td>1.176653</td>\n",
       "      <td>0.712117</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>-0.826526</td>\n",
       "      <td>-0.433602</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.751357</td>\n",
       "      <td>-1.291709</td>\n",
       "      <td>-1.667046</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>-0.732916</td>\n",
       "      <td>-0.506476</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.802851</td>\n",
       "      <td>-0.996217</td>\n",
       "      <td>-1.200860</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>-0.904097</td>\n",
       "      <td>-0.007621</td>\n",
       "      <td>B</td>\n",
       "      <td>0.133604</td>\n",
       "      <td>-0.439558</td>\n",
       "      <td>-1.875238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>-0.786333</td>\n",
       "      <td>-0.395776</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.884828</td>\n",
       "      <td>-1.261318</td>\n",
       "      <td>-1.443992</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>-0.533035</td>\n",
       "      <td>-0.396734</td>\n",
       "      <td>B</td>\n",
       "      <td>0.144547</td>\n",
       "      <td>0.030226</td>\n",
       "      <td>-0.227741</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-0.604797</td>\n",
       "      <td>-0.257585</td>\n",
       "      <td>B</td>\n",
       "      <td>0.406943</td>\n",
       "      <td>-0.411334</td>\n",
       "      <td>-0.528377</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>-0.505759</td>\n",
       "      <td>-0.169518</td>\n",
       "      <td>B</td>\n",
       "      <td>0.504564</td>\n",
       "      <td>0.243934</td>\n",
       "      <td>-0.082472</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>-0.973726</td>\n",
       "      <td>-0.294057</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.581649</td>\n",
       "      <td>-1.034058</td>\n",
       "      <td>-2.407620</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>-0.658269</td>\n",
       "      <td>-0.513102</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.816443</td>\n",
       "      <td>-0.655665</td>\n",
       "      <td>-0.828403</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>-0.781145</td>\n",
       "      <td>-0.319219</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.819085</td>\n",
       "      <td>-1.226320</td>\n",
       "      <td>-1.391629</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           x         y label    pred_q    pred_g    pred_p\n",
       "0  -0.103898  0.179248     A  1.667739  0.819567  1.309670\n",
       "1  -0.530864 -0.457307     A -0.155607  0.073109 -0.222972\n",
       "2  -0.210400 -0.424927     A  0.926252  1.369297  1.066187\n",
       "3  -0.021150 -0.468933     A  0.791680  1.234316  1.707608\n",
       "4  -0.208184 -0.487275     A  1.099681  1.354076  1.088123\n",
       "5   0.157078 -0.476244     A  0.049546  1.044874  2.214529\n",
       "6   0.116275  0.720881     A  0.342349  0.542134  1.735123\n",
       "7  -0.068440 -0.523605     A  1.166646  1.272920  1.579304\n",
       "8  -0.544330 -0.513921     A -0.489345  0.028090 -0.287809\n",
       "9  -0.310342 -0.539510     A  0.671727  1.176653  0.712117\n",
       "10 -0.826526 -0.433602     B -0.751357 -1.291709 -1.667046\n",
       "11 -0.732916 -0.506476     B -0.802851 -0.996217 -1.200860\n",
       "12 -0.904097 -0.007621     B  0.133604 -0.439558 -1.875238\n",
       "13 -0.786333 -0.395776     B -0.884828 -1.261318 -1.443992\n",
       "14 -0.533035 -0.396734     B  0.144547  0.030226 -0.227741\n",
       "15 -0.604797 -0.257585     B  0.406943 -0.411334 -0.528377\n",
       "16 -0.505759 -0.169518     B  0.504564  0.243934 -0.082472\n",
       "17 -0.973726 -0.294057     B -0.581649 -1.034058 -2.407620\n",
       "18 -0.658269 -0.513102     B -0.816443 -0.655665 -0.828403\n",
       "19 -0.781145 -0.319219     B -0.819085 -1.226320 -1.391629"
      ]
     },
     "execution_count": 51,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "records = []\n",
    "for label, data in test_input_unnormalized_br.items():\n",
    "    for x, y in data:\n",
    "        records.append({\"x\":x, \"y\":y, \"label\":label})\n",
    "df_test_br = pd.DataFrame(records)\n",
    "\n",
    "\n",
    "kernel_dict = {\"gaussian\":{\"mat\":g_kernel_br, \"func\":gaussian_kernel},\n",
    "                         \"polynominal\": {\"mat\":poly_kernel_br, \"func\":polynomial_kernel},\n",
    "                         \"quantum\": {\"mat\": q_kernel_br, \"func\": quantum_kernel}\n",
    "                        }\n",
    "\n",
    "# quantum\n",
    "kernel_choice = kernel_dict[\"quantum\"]\n",
    "data = df_train_br.iloc[:,:2].values\n",
    "label = df_train_br.iloc[:,2].apply(lambda x: 1.0 if x==\"A\" else -1.0)\n",
    "\n",
    "a, b = get_laglange_param(X=data,t=label, kernel=kernel_choice[\"mat\"])\n",
    "test_pred_q = array([f(array(row), a, label, data, b, kernel_choice[\"func\"]) for row in tqdm(df_test_br.iloc[:,:2].values,total=len(df_test_br))])\n",
    "\n",
    "\n",
    "df_test_br[\"pred_q\"] = test_pred_q\n",
    "\n",
    "# gaussian\n",
    "kernel_choice = kernel_dict[\"gaussian\"]\n",
    "data = df_train_br.iloc[:,:2].values\n",
    "label = df_train_br.iloc[:,2].apply(lambda x: 1.0 if x==\"A\" else -1.0)\n",
    "\n",
    "a, b = get_laglange_param(X=data,t=label, kernel=kernel_choice[\"mat\"])\n",
    "test_pred_g = array([f(array(row), a, label, data, b, kernel_choice[\"func\"]) for row in tqdm(df_test_br.iloc[:,:2].values,total=len(df_test_br))])\n",
    "\n",
    "\n",
    "df_test_br[\"pred_g\"] = test_pred_g\n",
    "\n",
    "# polynominal\n",
    "kernel_choice = kernel_dict[\"polynominal\"]\n",
    "data = df_train_br.iloc[:,:2].values\n",
    "label = df_train_br.iloc[:,2].apply(lambda x: 1.0 if x==\"A\" else -1.0)\n",
    "\n",
    "a, b = get_laglange_param(X=data,t=label, kernel=kernel_choice[\"mat\"])\n",
    "test_pred_p = array([f(array(row), a, label, data, b, kernel_choice[\"func\"]) for row in tqdm(df_test_br.iloc[:,:2].values,total=len(df_test_br))])\n",
    "\n",
    "\n",
    "df_test_br[\"pred_p\"] = test_pred_p\n",
    "\n",
    "df_test_br"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 52,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "breast cancer/Accuracy:  0.75\n",
      "Ad hoc/Accuracy:  1.0\n"
     ]
    }
   ],
   "source": [
    "# Comparison against Library\n",
    "\n",
    "os.environ['QISKIT_IN_PARALLEL'] = 'TRUE'  # brokeProcessError can be fix by this command\n",
    "shots=1024\n",
    "# Wall time: 9.47 s for train_size = 20 test_size = 10\n",
    "backend = BasicAer.get_backend('qasm_simulator')\n",
    "feature_map = ZZFeatureMap(feature_dim, reps=2)\n",
    "svm = QSVM(feature_map, training_input_unnormalized_br, test_input_unnormalized_br, None)  # the data for prediction can be fed later.\n",
    "svm.random_seed = seed\n",
    "quantum_instance = QuantumInstance(backend, shots=shots, seed_simulator=seed, seed_transpiler=seed)\n",
    "result = svm.run(quantum_instance)\n",
    "kernel_matrix = result['kernel_matrix_training']\n",
    "print(\"breast cancer/Accuracy: \", result[\"testing_accuracy\"])\n",
    "\n",
    "svm = QSVM(feature_map, training_input_unnormalized, test_input_unnormalized, None)  # the data for prediction can be fed later.\n",
    "svm.random_seed = seed\n",
    "quantum_instance = QuantumInstance(backend, shots=shots, seed_simulator=seed, seed_transpiler=seed)\n",
    "result = svm.run(quantum_instance)\n",
    "kernel_matrix = result['kernel_matrix_training']\n",
    "print(\"Ad hoc/Accuracy: \", result[\"testing_accuracy\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>label</th>\n",
       "      <th>pred_q</th>\n",
       "      <th>pred_g</th>\n",
       "      <th>pred_p</th>\n",
       "      <th>pred_q_lib</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.103898</td>\n",
       "      <td>0.179248</td>\n",
       "      <td>A</td>\n",
       "      <td>1.667739</td>\n",
       "      <td>0.819567</td>\n",
       "      <td>1.309670</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>-0.530864</td>\n",
       "      <td>-0.457307</td>\n",
       "      <td>A</td>\n",
       "      <td>-0.155607</td>\n",
       "      <td>0.073109</td>\n",
       "      <td>-0.222972</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>-0.210400</td>\n",
       "      <td>-0.424927</td>\n",
       "      <td>A</td>\n",
       "      <td>0.926252</td>\n",
       "      <td>1.369297</td>\n",
       "      <td>1.066187</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.021150</td>\n",
       "      <td>-0.468933</td>\n",
       "      <td>A</td>\n",
       "      <td>0.791680</td>\n",
       "      <td>1.234316</td>\n",
       "      <td>1.707608</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>-0.208184</td>\n",
       "      <td>-0.487275</td>\n",
       "      <td>A</td>\n",
       "      <td>1.099681</td>\n",
       "      <td>1.354076</td>\n",
       "      <td>1.088123</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>5</th>\n",
       "      <td>0.157078</td>\n",
       "      <td>-0.476244</td>\n",
       "      <td>A</td>\n",
       "      <td>0.049546</td>\n",
       "      <td>1.044874</td>\n",
       "      <td>2.214529</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6</th>\n",
       "      <td>0.116275</td>\n",
       "      <td>0.720881</td>\n",
       "      <td>A</td>\n",
       "      <td>0.342349</td>\n",
       "      <td>0.542134</td>\n",
       "      <td>1.735123</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>-0.068440</td>\n",
       "      <td>-0.523605</td>\n",
       "      <td>A</td>\n",
       "      <td>1.166646</td>\n",
       "      <td>1.272920</td>\n",
       "      <td>1.579304</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>-0.544330</td>\n",
       "      <td>-0.513921</td>\n",
       "      <td>A</td>\n",
       "      <td>-0.489345</td>\n",
       "      <td>0.028090</td>\n",
       "      <td>-0.287809</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>-0.310342</td>\n",
       "      <td>-0.539510</td>\n",
       "      <td>A</td>\n",
       "      <td>0.671727</td>\n",
       "      <td>1.176653</td>\n",
       "      <td>0.712117</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>10</th>\n",
       "      <td>-0.826526</td>\n",
       "      <td>-0.433602</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.751357</td>\n",
       "      <td>-1.291709</td>\n",
       "      <td>-1.667046</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>11</th>\n",
       "      <td>-0.732916</td>\n",
       "      <td>-0.506476</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.802851</td>\n",
       "      <td>-0.996217</td>\n",
       "      <td>-1.200860</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>12</th>\n",
       "      <td>-0.904097</td>\n",
       "      <td>-0.007621</td>\n",
       "      <td>B</td>\n",
       "      <td>0.133604</td>\n",
       "      <td>-0.439558</td>\n",
       "      <td>-1.875238</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>13</th>\n",
       "      <td>-0.786333</td>\n",
       "      <td>-0.395776</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.884828</td>\n",
       "      <td>-1.261318</td>\n",
       "      <td>-1.443992</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>14</th>\n",
       "      <td>-0.533035</td>\n",
       "      <td>-0.396734</td>\n",
       "      <td>B</td>\n",
       "      <td>0.144547</td>\n",
       "      <td>0.030226</td>\n",
       "      <td>-0.227741</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>15</th>\n",
       "      <td>-0.604797</td>\n",
       "      <td>-0.257585</td>\n",
       "      <td>B</td>\n",
       "      <td>0.406943</td>\n",
       "      <td>-0.411334</td>\n",
       "      <td>-0.528377</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>16</th>\n",
       "      <td>-0.505759</td>\n",
       "      <td>-0.169518</td>\n",
       "      <td>B</td>\n",
       "      <td>0.504564</td>\n",
       "      <td>0.243934</td>\n",
       "      <td>-0.082472</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>-0.973726</td>\n",
       "      <td>-0.294057</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.581649</td>\n",
       "      <td>-1.034058</td>\n",
       "      <td>-2.407620</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>-0.658269</td>\n",
       "      <td>-0.513102</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.816443</td>\n",
       "      <td>-0.655665</td>\n",
       "      <td>-0.828403</td>\n",
       "      <td>0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>19</th>\n",
       "      <td>-0.781145</td>\n",
       "      <td>-0.319219</td>\n",
       "      <td>B</td>\n",
       "      <td>-0.819085</td>\n",
       "      <td>-1.226320</td>\n",
       "      <td>-1.391629</td>\n",
       "      <td>1</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           x         y label    pred_q    pred_g    pred_p  pred_q_lib\n",
       "0  -0.103898  0.179248     A  1.667739  0.819567  1.309670           0\n",
       "1  -0.530864 -0.457307     A -0.155607  0.073109 -0.222972           1\n",
       "2  -0.210400 -0.424927     A  0.926252  1.369297  1.066187           0\n",
       "3  -0.021150 -0.468933     A  0.791680  1.234316  1.707608           0\n",
       "4  -0.208184 -0.487275     A  1.099681  1.354076  1.088123           0\n",
       "5   0.157078 -0.476244     A  0.049546  1.044874  2.214529           0\n",
       "6   0.116275  0.720881     A  0.342349  0.542134  1.735123           1\n",
       "7  -0.068440 -0.523605     A  1.166646  1.272920  1.579304           0\n",
       "8  -0.544330 -0.513921     A -0.489345  0.028090 -0.287809           1\n",
       "9  -0.310342 -0.539510     A  0.671727  1.176653  0.712117           1\n",
       "10 -0.826526 -0.433602     B -0.751357 -1.291709 -1.667046           1\n",
       "11 -0.732916 -0.506476     B -0.802851 -0.996217 -1.200860           0\n",
       "12 -0.904097 -0.007621     B  0.133604 -0.439558 -1.875238           1\n",
       "13 -0.786333 -0.395776     B -0.884828 -1.261318 -1.443992           1\n",
       "14 -0.533035 -0.396734     B  0.144547  0.030226 -0.227741           1\n",
       "15 -0.604797 -0.257585     B  0.406943 -0.411334 -0.528377           1\n",
       "16 -0.505759 -0.169518     B  0.504564  0.243934 -0.082472           1\n",
       "17 -0.973726 -0.294057     B -0.581649 -1.034058 -2.407620           1\n",
       "18 -0.658269 -0.513102     B -0.816443 -0.655665 -0.828403           0\n",
       "19 -0.781145 -0.319219     B -0.819085 -1.226320 -1.391629           1"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# breast Cancer Result\n",
    "pred_q_lib = svm.predict(df_test_br.iloc[:,:2].values)\n",
    "\n",
    "df_test_br[\"pred_q_lib\"] = pred_q_lib\n",
    "df_test_br"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "breast cancer/Accuracy:  0.7\n",
      "breast cancer/Accuracy:  1.0\n"
     ]
    }
   ],
   "source": [
    "# Accuracy of QSVM scractch\n",
    "# breast Cancer accuracy \n",
    "print(\"breast cancer/Accuracy: \",\n",
    "         df_test_br.apply(lambda x: (x[\"label\"] ==\"A\" and x[\"pred_q\"] >0) or (x[\"label\"] ==\"B\" and x[\"pred_q\"] < 0), axis=1).sum()/len(df_test_br))\n",
    "# ad_hoc accuracy \n",
    "print(\"breast cancer/Accuracy: \",\n",
    "          df_test.apply(lambda x: (x[\"label\"] ==\"A\" and x[\"pred_q\"] >0) or (x[\"label\"] ==\"B\" and x[\"pred_q\"] < 0), axis=1).sum()/len(df_test))"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# Periodic Data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": [
    "# I expect QSVM can classify periodic boundary well\n",
    "# It turned out to be not the case..."
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 56,
   "metadata": {},
   "outputs": [],
   "source": [
    "import random\n",
    "periodic_data = np.zeros((100,100))\n",
    "flip_lists = sum([list(range(i*10, i*10+10)) for i in  range(0,10, 2)],[])\n",
    "periodic_data[: , flip_lists] = 1\n",
    "\n",
    "training_data_num = 40\n",
    "test_data_num = 40\n",
    "\n",
    "random.seed(1)\n",
    "random_ind_x = random.sample(range(100),training_data_num + test_data_num)\n",
    "random_ind_y = random.sample(range(100),training_data_num + test_data_num)\n",
    "traing_data_label = periodic_data[random_ind_x, random_ind_y]\n",
    "df_period = pd.DataFrame({\"x\":np.array(random_ind_x)/100*2*np.pi, \"y\": np.array(random_ind_y)/100*2*np.pi, \"label\": traing_data_label})\n",
    "df_period_train = df_period[:training_data_num].copy()\n",
    "df_period_test = df_period[training_data_num:].copy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 57,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.axes._subplots.AxesSubplot at 0x7fdf627a94d0>"
      ]
     },
     "execution_count": 57,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 648x432 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(9, 6))\n",
    "sns.heatmap(periodic_data.T, annot=False, fmt='g', cmap='Blues').invert_yaxis()\n",
    "\n",
    "df_plot = df_period_train.copy()\n",
    "df_plot[\"x\"] = df_plot[\"x\"]*100/2/np.pi\n",
    "df_plot[\"y\"] = df_plot[\"y\"]*100/2/np.pi\n",
    "sns.scatterplot(data=df_plot, x=\"x\", y=\"y\", hue=\"label\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 58,
   "metadata": {},
   "outputs": [],
   "source": [
    "training_input_pr = {\"A\": df_period_train.query(\"label == 0\").iloc[:,:2].values, \"B\": df_period_train.query(\"label == 0\").iloc[:,:2].values}\n",
    "test_input_pr = {\"A\": df_period_test.query(\"label == 0\").iloc[:,:2].values,  \"B\": df_period_test.query(\"label == 0\").iloc[:,:2].values}"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Gaussian Ver"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 59,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "array([0., 1., 1., 0., 1., 0., 0., 1., 1., 0., 0., 0., 0., 0., 1., 1., 0.,\n",
       "       1., 0., 0.])"
      ]
     },
     "execution_count": 59,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from qiskit.aqua.algorithms import SklearnSVM\n",
    "\n",
    "result = SklearnSVM(training_input_pr, test_input_pr,df_test.iloc[:,:2].values,  gamma=0.7).run()\n",
    "\n",
    "print(result[\"testing_accuracy\"])\n",
    "result[\"predicted_labels\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 60,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array([[1.00000000e+000, 7.90721570e-049, 1.90365327e-112, ...,\n",
       "        2.98824754e-006, 2.75230052e-057, 6.46988043e-059],\n",
       "       [7.90721570e-049, 1.00000000e+000, 3.79378158e-014, ...,\n",
       "        8.90086274e-022, 1.59833869e-042, 6.52016735e-035],\n",
       "       [1.90365327e-112, 3.79378158e-014, 1.00000000e+000, ...,\n",
       "        2.74255100e-068, 1.69387655e-066, 1.52383267e-054],\n",
       "       ...,\n",
       "       [2.98824754e-006, 8.90086274e-022, 2.74255100e-068, ...,\n",
       "        1.00000000e+000, 2.27170179e-040, 5.96482563e-039],\n",
       "       [2.75230052e-057, 1.59833869e-042, 1.69387655e-066, ...,\n",
       "        2.27170179e-040, 1.00000000e+000, 2.40361912e-001],\n",
       "       [6.46988043e-059, 6.52016735e-035, 1.52383267e-054, ...,\n",
       "        5.96482563e-039, 2.40361912e-001, 1.00000000e+000]])"
      ]
     },
     "execution_count": 60,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "sigma=0.3\n",
    "g_kernel_pr = [] \n",
    "for i in range(len(df_period_train)):\n",
    "    g_kernel_pr.append(\n",
    "            [gaussian_kernel(df_period_train.iloc[i,:2], df_period_train.iloc[ii,:2]) for ii in range(len(df_period_train))]\n",
    "    )\n",
    "g_kernel_pr = np.array(g_kernel_pr)  \n",
    "g_kernel_pr "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 62,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -1.4693e+01 -4.0899e+01  2e+02  4e+00  3e-16\n",
      " 1: -1.0910e+01 -3.1524e+01  2e+01  1e-15  4e-16\n",
      " 2: -1.1818e+01 -1.3636e+01  2e+00  2e-16  2e-16\n",
      " 3: -1.2141e+01 -1.2299e+01  2e-01  4e-16  3e-16\n",
      " 4: -1.2176e+01 -1.2182e+01  5e-03  7e-16  3e-16\n",
      " 5: -1.2177e+01 -1.2178e+01  1e-04  7e-16  3e-16\n",
      " 6: -1.2177e+01 -1.2177e+01  1e-06  4e-16  3e-16\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[0.5        0.5        0.1904322  0.29812014 0.5        0.29217291\n",
      " 0.33289839 0.5        0.49999988 0.42541212 0.29445723 0.49999973\n",
      " 0.42066555 0.5        0.5        0.5        0.5        0.5\n",
      " 0.28686363 0.2131368  0.20880268 0.20339695 0.29123546 0.13580011\n",
      " 0.27665264 0.5        0.45042461 0.5        0.5        0.29122567\n",
      " 0.29251611 0.5        0.5        0.29123412 0.34106211 0.5\n",
      " 0.2911617  0.21226341 0.22229713 0.23776869]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "f9c55a6afa814f0f81c66bd4e8f0d18c",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=900.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 432x288 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "24f7cbb1ca6e43599ed3b13c8dbee84b",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=40.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>label</th>\n",
       "      <th>pred_g</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>3.832743</td>\n",
       "      <td>1.884956</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.056688</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>4.021239</td>\n",
       "      <td>4.775221</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.217949</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>2.450442</td>\n",
       "      <td>5.717699</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.262947</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>4.963716</td>\n",
       "      <td>5.654867</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.181177</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>1.445133</td>\n",
       "      <td>3.330088</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.339345</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>4.586725</td>\n",
       "      <td>4.963716</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.250723</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>2.953097</td>\n",
       "      <td>2.324779</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.120392</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>2.827433</td>\n",
       "      <td>6.220353</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.262760</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>1.193805</td>\n",
       "      <td>0.879646</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.002956</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>5.466371</td>\n",
       "      <td>4.838053</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.202125</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>3.267256</td>\n",
       "      <td>3.581416</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.181143</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>4.272566</td>\n",
       "      <td>4.586725</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.315091</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>2.638938</td>\n",
       "      <td>2.387610</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.253913</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>1.130973</td>\n",
       "      <td>0.753982</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.096296</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>2.701770</td>\n",
       "      <td>3.455752</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.268205</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>0.376991</td>\n",
       "      <td>3.267256</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.310832</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>5.403539</td>\n",
       "      <td>1.947787</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.323468</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>4.461062</td>\n",
       "      <td>2.261947</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.191895</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>4.398230</td>\n",
       "      <td>2.010619</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.219393</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>4.900885</td>\n",
       "      <td>2.199115</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.262849</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>5.152212</td>\n",
       "      <td>5.843362</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.028774</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>5.717699</td>\n",
       "      <td>0.439823</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.297286</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>6.220353</td>\n",
       "      <td>0.502655</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.332872</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>0.314159</td>\n",
       "      <td>1.507964</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.454001</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>3.644247</td>\n",
       "      <td>1.759292</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.058847</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>4.775221</td>\n",
       "      <td>2.513274</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.227358</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>3.707079</td>\n",
       "      <td>1.445133</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.100810</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>5.906194</td>\n",
       "      <td>0.251327</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.407213</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>2.890265</td>\n",
       "      <td>4.209734</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.255890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>1.570796</td>\n",
       "      <td>5.466371</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.456855</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>1.947787</td>\n",
       "      <td>4.272566</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.460928</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>5.529203</td>\n",
       "      <td>3.958407</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.362610</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>2.576106</td>\n",
       "      <td>1.382301</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.304975</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>0.691150</td>\n",
       "      <td>5.152212</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.327840</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>5.969026</td>\n",
       "      <td>5.026548</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.550510</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>1.319469</td>\n",
       "      <td>5.215044</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.412162</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>0.439823</td>\n",
       "      <td>5.403539</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.501879</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>4.209734</td>\n",
       "      <td>2.890265</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.160478</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>1.884956</td>\n",
       "      <td>0.125664</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.228970</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>1.507964</td>\n",
       "      <td>3.832743</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.404255</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           x         y  label    pred_g\n",
       "40  3.832743  1.884956    0.0  0.056688\n",
       "75  4.021239  4.775221    0.0  0.217949\n",
       "71  2.450442  5.717699    0.0  0.262947\n",
       "70  4.963716  5.654867    0.0 -0.181177\n",
       "69  1.445133  3.330088    0.0  0.339345\n",
       "66  4.586725  4.963716    0.0  0.250723\n",
       "64  2.953097  2.324779    0.0  0.120392\n",
       "62  2.827433  6.220353    0.0  0.262760\n",
       "61  1.193805  0.879646    0.0 -0.002956\n",
       "60  5.466371  4.838053    0.0  0.202125\n",
       "78  3.267256  3.581416    0.0  0.181143\n",
       "58  4.272566  4.586725    0.0  0.315091\n",
       "59  2.638938  2.387610    0.0  0.253913\n",
       "45  1.130973  0.753982    0.0 -0.096296\n",
       "41  2.701770  3.455752    0.0  0.268205\n",
       "50  0.376991  3.267256    0.0  0.310832\n",
       "48  5.403539  1.947787    0.0  0.323468\n",
       "43  4.461062  2.261947    0.0  0.191895\n",
       "46  4.398230  2.010619    0.0  0.219393\n",
       "44  4.900885  2.199115    0.0  0.262849\n",
       "52  5.152212  5.843362    0.0 -0.028774\n",
       "77  5.717699  0.439823    1.0  0.297286\n",
       "76  6.220353  0.502655    1.0  0.332872\n",
       "74  0.314159  1.507964    1.0  0.454001\n",
       "73  3.644247  1.759292    1.0  0.058847\n",
       "72  4.775221  2.513274    1.0  0.227358\n",
       "42  3.707079  1.445133    1.0  0.100810\n",
       "68  5.906194  0.251327    1.0  0.407213\n",
       "53  2.890265  4.209734    1.0  0.255890\n",
       "65  1.570796  5.466371    1.0  0.456855\n",
       "63  1.947787  4.272566    1.0  0.460928\n",
       "47  5.529203  3.958407    1.0  0.362610\n",
       "49  2.576106  1.382301    1.0  0.304975\n",
       "51  0.691150  5.152212    1.0  0.327840\n",
       "57  5.969026  5.026548    1.0  0.550510\n",
       "56  1.319469  5.215044    1.0  0.412162\n",
       "55  0.439823  5.403539    1.0  0.501879\n",
       "54  4.209734  2.890265    1.0  0.160478\n",
       "67  1.884956  0.125664    1.0  0.228970\n",
       "79  1.507964  3.832743    1.0  0.404255"
      ]
     },
     "execution_count": 62,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C = 0.5        # hyper-parameter of soft-margin SVM\n",
    "\n",
    "\n",
    "# load traing dataset as input\n",
    "data = df_period_train.iloc[:,0:2].values\n",
    "label = df_period_train[\"label\"].apply(lambda x: x*2-1)\n",
    " \n",
    "#  decide the choice of kernel\n",
    "kernel_choice = kernel_dict[\"gaussian\"]\n",
    "\n",
    "a, b = get_laglange_param(t=label, kernel=g_kernel_pr)\n",
    "draw_contour(X=data,t=label, kernel_func=kernel_choice[\"func\"])\n",
    "\n",
    "test_pred_g = array([f(array(row), a, label, data, b, kernel_choice[\"func\"]) for row in tqdm(df_period_test.iloc[:,:2].values,total=len(df_period_test))])\n",
    "\n",
    "\n",
    "df_period_test[\"pred_g\"] = test_pred_g\n",
    "df_period_test.sort_values(\"label\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 63,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.575"
      ]
     },
     "execution_count": 63,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Accuracy\n",
    "df_period_test.apply(lambda x: (x[\"label\"]==1 and x[\"pred_g\"]>0) or (x[\"label\"]==0 and x[\"pred_g\"]<0),axis=1).sum()/len(df_period_test)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Quantume Ver"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Scratch"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 64,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "1191cc8f63c945a48664dc4514417ced",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=40.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    }
   ],
   "source": [
    "q_kernel_pr = []\n",
    "for ind, row in tqdm(df_period_train.iterrows(),total=len(df_period_train)):\n",
    "    q_kernel_pr.append([quantum_kernel(row[:2].values, row_in[:2].values) for ind_in, row_in in df_period_train.iterrows()])\n",
    "q_kernel_pr =  np.array(q_kernel_pr)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 65,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "     pcost       dcost       gap    pres   dres\n",
      " 0: -2.2263e+01 -4.7696e+01  3e+02  5e+00  4e-16\n",
      " 1: -1.1932e+01 -3.8638e+01  3e+01  7e-16  5e-16\n",
      " 2: -1.3325e+01 -1.5805e+01  2e+00  2e-16  5e-16\n",
      " 3: -1.4025e+01 -1.4507e+01  5e-01  1e-16  4e-16\n",
      " 4: -1.4205e+01 -1.4307e+01  1e-01  2e-16  4e-16\n",
      " 5: -1.4238e+01 -1.4269e+01  3e-02  4e-16  5e-16\n",
      " 6: -1.4251e+01 -1.4255e+01  4e-03  2e-16  4e-16\n",
      " 7: -1.4253e+01 -1.4253e+01  7e-05  9e-16  4e-16\n",
      " 8: -1.4253e+01 -1.4253e+01  7e-07  4e-16  4e-16\n",
      "Optimal solution found.\n",
      "Lagrange Param\n",
      "[4.99999997e-01 4.99999997e-01 4.99998454e-01 3.14306906e-07\n",
      " 4.99999995e-01 1.03811481e-07 4.99999961e-01 4.99999997e-01\n",
      " 7.87229481e-08 1.05246256e-01 4.68250770e-08 4.99999965e-01\n",
      " 2.62786683e-01 4.99999996e-01 4.99999994e-01 4.99999997e-01\n",
      " 4.99999996e-01 4.99999997e-01 4.99999802e-01 4.99999885e-01\n",
      " 3.29576499e-01 9.06966943e-08 4.99999896e-01 4.59669410e-02\n",
      " 2.53552138e-07 4.99999996e-01 4.99999869e-01 4.99999996e-01\n",
      " 4.99999996e-01 4.99999978e-01 2.83502923e-01 4.99999995e-01\n",
      " 4.99999997e-01 3.10648728e-08 4.99999988e-01 4.99999996e-01\n",
      " 4.99998971e-01 4.72927250e-01 4.99999973e-01 4.99995729e-01]\n"
     ]
    },
    {
     "data": {
      "application/vnd.jupyter.widget-view+json": {
       "model_id": "624bdd5688374352b9902280ba9dfc66",
       "version_major": 2,
       "version_minor": 0
      },
      "text/plain": [
       "HBox(children=(FloatProgress(value=0.0, max=40.0), HTML(value='')))"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n"
     ]
    },
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>x</th>\n",
       "      <th>y</th>\n",
       "      <th>label</th>\n",
       "      <th>pred_g</th>\n",
       "      <th>pred_q</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>40</th>\n",
       "      <td>3.832743</td>\n",
       "      <td>1.884956</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.056688</td>\n",
       "      <td>0.200710</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>75</th>\n",
       "      <td>4.021239</td>\n",
       "      <td>4.775221</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.217949</td>\n",
       "      <td>0.479167</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>71</th>\n",
       "      <td>2.450442</td>\n",
       "      <td>5.717699</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.262947</td>\n",
       "      <td>0.041888</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>70</th>\n",
       "      <td>4.963716</td>\n",
       "      <td>5.654867</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.181177</td>\n",
       "      <td>0.544535</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>69</th>\n",
       "      <td>1.445133</td>\n",
       "      <td>3.330088</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.339345</td>\n",
       "      <td>-0.288935</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>66</th>\n",
       "      <td>4.586725</td>\n",
       "      <td>4.963716</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.250723</td>\n",
       "      <td>0.349674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>64</th>\n",
       "      <td>2.953097</td>\n",
       "      <td>2.324779</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.120392</td>\n",
       "      <td>0.515607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>62</th>\n",
       "      <td>2.827433</td>\n",
       "      <td>6.220353</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.262760</td>\n",
       "      <td>0.423320</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>61</th>\n",
       "      <td>1.193805</td>\n",
       "      <td>0.879646</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.002956</td>\n",
       "      <td>0.475110</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>60</th>\n",
       "      <td>5.466371</td>\n",
       "      <td>4.838053</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.202125</td>\n",
       "      <td>0.230873</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>78</th>\n",
       "      <td>3.267256</td>\n",
       "      <td>3.581416</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.181143</td>\n",
       "      <td>0.559131</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>58</th>\n",
       "      <td>4.272566</td>\n",
       "      <td>4.586725</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.315091</td>\n",
       "      <td>0.659105</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>59</th>\n",
       "      <td>2.638938</td>\n",
       "      <td>2.387610</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.253913</td>\n",
       "      <td>0.312669</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>45</th>\n",
       "      <td>1.130973</td>\n",
       "      <td>0.753982</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.096296</td>\n",
       "      <td>0.393652</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>41</th>\n",
       "      <td>2.701770</td>\n",
       "      <td>3.455752</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.268205</td>\n",
       "      <td>0.570830</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>50</th>\n",
       "      <td>0.376991</td>\n",
       "      <td>3.267256</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.310832</td>\n",
       "      <td>0.433780</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>48</th>\n",
       "      <td>5.403539</td>\n",
       "      <td>1.947787</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.323468</td>\n",
       "      <td>0.159543</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>43</th>\n",
       "      <td>4.461062</td>\n",
       "      <td>2.261947</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.191895</td>\n",
       "      <td>0.521773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>46</th>\n",
       "      <td>4.398230</td>\n",
       "      <td>2.010619</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.219393</td>\n",
       "      <td>0.677626</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>44</th>\n",
       "      <td>4.900885</td>\n",
       "      <td>2.199115</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.262849</td>\n",
       "      <td>0.566773</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>52</th>\n",
       "      <td>5.152212</td>\n",
       "      <td>5.843362</td>\n",
       "      <td>0.0</td>\n",
       "      <td>-0.028774</td>\n",
       "      <td>0.351035</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>77</th>\n",
       "      <td>5.717699</td>\n",
       "      <td>0.439823</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.297286</td>\n",
       "      <td>0.188238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>76</th>\n",
       "      <td>6.220353</td>\n",
       "      <td>0.502655</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.332872</td>\n",
       "      <td>0.236305</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>74</th>\n",
       "      <td>0.314159</td>\n",
       "      <td>1.507964</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.454001</td>\n",
       "      <td>-0.205722</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>73</th>\n",
       "      <td>3.644247</td>\n",
       "      <td>1.759292</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.058847</td>\n",
       "      <td>-0.040136</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>72</th>\n",
       "      <td>4.775221</td>\n",
       "      <td>2.513274</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.227358</td>\n",
       "      <td>0.399238</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>42</th>\n",
       "      <td>3.707079</td>\n",
       "      <td>1.445133</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.100810</td>\n",
       "      <td>0.072703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>68</th>\n",
       "      <td>5.906194</td>\n",
       "      <td>0.251327</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.407213</td>\n",
       "      <td>0.134418</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>53</th>\n",
       "      <td>2.890265</td>\n",
       "      <td>4.209734</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.255890</td>\n",
       "      <td>0.400927</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>65</th>\n",
       "      <td>1.570796</td>\n",
       "      <td>5.466371</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.456855</td>\n",
       "      <td>0.165890</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>63</th>\n",
       "      <td>1.947787</td>\n",
       "      <td>4.272566</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.460928</td>\n",
       "      <td>0.738552</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>47</th>\n",
       "      <td>5.529203</td>\n",
       "      <td>3.958407</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.362610</td>\n",
       "      <td>0.490930</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>49</th>\n",
       "      <td>2.576106</td>\n",
       "      <td>1.382301</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.304975</td>\n",
       "      <td>0.094436</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>51</th>\n",
       "      <td>0.691150</td>\n",
       "      <td>5.152212</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.327840</td>\n",
       "      <td>-0.093674</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>57</th>\n",
       "      <td>5.969026</td>\n",
       "      <td>5.026548</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.550510</td>\n",
       "      <td>0.360077</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>56</th>\n",
       "      <td>1.319469</td>\n",
       "      <td>5.215044</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.412162</td>\n",
       "      <td>0.181598</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>55</th>\n",
       "      <td>0.439823</td>\n",
       "      <td>5.403539</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.501879</td>\n",
       "      <td>0.589536</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>54</th>\n",
       "      <td>4.209734</td>\n",
       "      <td>2.890265</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.160478</td>\n",
       "      <td>0.069361</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>67</th>\n",
       "      <td>1.884956</td>\n",
       "      <td>0.125664</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.228970</td>\n",
       "      <td>0.169784</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>79</th>\n",
       "      <td>1.507964</td>\n",
       "      <td>3.832743</td>\n",
       "      <td>1.0</td>\n",
       "      <td>0.404255</td>\n",
       "      <td>0.390372</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "           x         y  label    pred_g    pred_q\n",
       "40  3.832743  1.884956    0.0  0.056688  0.200710\n",
       "75  4.021239  4.775221    0.0  0.217949  0.479167\n",
       "71  2.450442  5.717699    0.0  0.262947  0.041888\n",
       "70  4.963716  5.654867    0.0 -0.181177  0.544535\n",
       "69  1.445133  3.330088    0.0  0.339345 -0.288935\n",
       "66  4.586725  4.963716    0.0  0.250723  0.349674\n",
       "64  2.953097  2.324779    0.0  0.120392  0.515607\n",
       "62  2.827433  6.220353    0.0  0.262760  0.423320\n",
       "61  1.193805  0.879646    0.0 -0.002956  0.475110\n",
       "60  5.466371  4.838053    0.0  0.202125  0.230873\n",
       "78  3.267256  3.581416    0.0  0.181143  0.559131\n",
       "58  4.272566  4.586725    0.0  0.315091  0.659105\n",
       "59  2.638938  2.387610    0.0  0.253913  0.312669\n",
       "45  1.130973  0.753982    0.0 -0.096296  0.393652\n",
       "41  2.701770  3.455752    0.0  0.268205  0.570830\n",
       "50  0.376991  3.267256    0.0  0.310832  0.433780\n",
       "48  5.403539  1.947787    0.0  0.323468  0.159543\n",
       "43  4.461062  2.261947    0.0  0.191895  0.521773\n",
       "46  4.398230  2.010619    0.0  0.219393  0.677626\n",
       "44  4.900885  2.199115    0.0  0.262849  0.566773\n",
       "52  5.152212  5.843362    0.0 -0.028774  0.351035\n",
       "77  5.717699  0.439823    1.0  0.297286  0.188238\n",
       "76  6.220353  0.502655    1.0  0.332872  0.236305\n",
       "74  0.314159  1.507964    1.0  0.454001 -0.205722\n",
       "73  3.644247  1.759292    1.0  0.058847 -0.040136\n",
       "72  4.775221  2.513274    1.0  0.227358  0.399238\n",
       "42  3.707079  1.445133    1.0  0.100810  0.072703\n",
       "68  5.906194  0.251327    1.0  0.407213  0.134418\n",
       "53  2.890265  4.209734    1.0  0.255890  0.400927\n",
       "65  1.570796  5.466371    1.0  0.456855  0.165890\n",
       "63  1.947787  4.272566    1.0  0.460928  0.738552\n",
       "47  5.529203  3.958407    1.0  0.362610  0.490930\n",
       "49  2.576106  1.382301    1.0  0.304975  0.094436\n",
       "51  0.691150  5.152212    1.0  0.327840 -0.093674\n",
       "57  5.969026  5.026548    1.0  0.550510  0.360077\n",
       "56  1.319469  5.215044    1.0  0.412162  0.181598\n",
       "55  0.439823  5.403539    1.0  0.501879  0.589536\n",
       "54  4.209734  2.890265    1.0  0.160478  0.069361\n",
       "67  1.884956  0.125664    1.0  0.228970  0.169784\n",
       "79  1.507964  3.832743    1.0  0.404255  0.390372"
      ]
     },
     "execution_count": 65,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "C = 0.5        # hyper-parameter of soft-margin SVM\n",
    "\n",
    "# load traing dataset as input\n",
    "data = df_period_train.iloc[:,0:2].values\n",
    "label = df_period_train[\"label\"].apply(lambda x: x*2-1)\n",
    " \n",
    "# decide the choice of kernel\n",
    "kernel_choice = kernel_dict[\"quantum\"]\n",
    "\n",
    "a, b = get_laglange_param(t=label, kernel=q_kernel_pr)\n",
    "# draw_contour(X=data,t=label, kernel_func=kernel_choice[\"func\"])\n",
    "test_pred_q = array([f(array(row), a, label, data, b, kernel_choice[\"func\"]) for row in tqdm(df_period_test.iloc[:,:2].values,total=len(df_period_test))])\n",
    "\n",
    "\n",
    "df_period_test[\"pred_q\"] = test_pred_q\n",
    "df_period_test.sort_values(\"label\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 66,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "0.425"
      ]
     },
     "execution_count": 66,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# Accuracy\n",
    "df_period_test.apply(lambda x: (x[\"label\"]==1 and x[\"pred_q\"]>0) or (x[\"label\"]==0 and x[\"pred_q\"]<0),axis=1).sum()/len(df_period_test)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## library 使用"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 67,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "0.5\n"
     ]
    },
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/y.n./.pyenv/versions/anaconda3-2020.02/lib/python3.7/site-packages/qiskit/aqua/algorithms/classifiers/qsvm/_qsvm_binary.py:109: ComplexWarning: Casting complex values to real discards the imaginary part\n",
      "  predicted_labels = binarized_predictions.astype(int)\n"
     ]
    }
   ],
   "source": [
    "feature_dim=2\n",
    "os.environ['QISKIT_IN_PARALLEL'] = 'TRUE'\n",
    "quantum_instance = QuantumInstance(backend, shots=1024, seed_simulator=seed, seed_transpiler=seed)\n",
    "feature_map = ZZFeatureMap(feature_dimension=feature_dim, reps=2, entanglement='linear')\n",
    "qsvm = QSVM(feature_map, training_input_pr, test_input_pr, lambda2=0.2)\n",
    "result = qsvm.run(quantum_instance)\n",
    "print(result[\"testing_accuracy\"])\n",
    "# qsvm.predict(split_dataset_to_data_and_labels(test_input_pr)[0][0])"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.7.6"
  },
  "toc": {
   "base_numbering": 1,
   "nav_menu": {},
   "number_sections": true,
   "sideBar": true,
   "skip_h1_title": false,
   "title_cell": "Table of Contents",
   "title_sidebar": "Contents",
   "toc_cell": false,
   "toc_position": {
    "height": "calc(100% - 180px)",
    "left": "10px",
    "top": "150px",
    "width": "358.391px"
   },
   "toc_section_display": true,
   "toc_window_display": true
  }
 },
 "nbformat": 4,
 "nbformat_minor": 4
}