1.2.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "id": "63372ade",
   "metadata": {},
   "source": [
    "# Imports"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "9780689b",
   "metadata": {},
   "outputs": [],
   "source": [
    "import pandas as pd\n",
    "import numpy as np\n",
    "from sklearn.preprocessing import *\n",
    "from sklearn.ensemble import RandomForestClassifier\n",
    "from sklearn.metrics import *\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "b86b0fa6",
   "metadata": {},
   "source": [
    "# Importing rpy2 and dependencies"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "c22fbe78",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "Unable to determine R home: [WinError 2] The system cannot find the file specified\n"
     ]
    }
   ],
   "source": [
    "import rpy2\n",
    "import rpy2.robjects as robjects\n",
    "from rpy2.robjects.packages import importr\n",
    "from rpy2.robjects import pandas2ri\n",
    "from rpy2.robjects.conversion import localconverter"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "1a9330f8",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "3.5.6\n"
     ]
    }
   ],
   "source": [
    "print(rpy2.__version__)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "2b5c7cfd",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<rpy2.rinterface_lib.sexp.NULLType object at 0x0000013B076BA240> [RTYPES.NILSXP]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# import rpy2's package module\n",
    "import rpy2.robjects.packages as rpackages\n",
    "\n",
    "# import R's utility package\n",
    "utils = rpackages.importr('utils')\n",
    "\n",
    "# select a mirror for R packages\n",
    "utils.chooseCRANmirror(ind=1) # select the first mirror in the list"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "8a083381",
   "metadata": {},
   "source": [
    "# Installing R packages"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "4da8df78",
   "metadata": {},
   "outputs": [],
   "source": [
    "# R package names\n",
    "packnames = (\"zCompositions\", \"robCompositions\")\n",
    "\n",
    "# R vector of strings\n",
    "from rpy2.robjects.vectors import StrVector\n",
    "\n",
    "# Selectively install what needs to be install.\n",
    "names_to_install = [x for x in packnames if not rpackages.isinstalled(x)]\n",
    "if len(names_to_install) > 0:\n",
    "    utils.install_packages(StrVector(names_to_install))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "9bb785fd",
   "metadata": {},
   "outputs": [],
   "source": [
    "zCompositions = importr('zCompositions')\n",
    "robCompositions = importr(\"robCompositions\")"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "52a45540",
   "metadata": {},
   "source": [
    "# Import data"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "458cb2d6",
   "metadata": {},
   "outputs": [],
   "source": [
    "path = 'CRC/'\n",
    "df = pd.read_csv(path + \"/all_data.csv\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "3dd0d308",
   "metadata": {},
   "outputs": [],
   "source": [
    "lbl=LabelEncoder()\n",
    "df['condition'] = lbl.fit_transform(df['condition'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "57ef2350",
   "metadata": {},
   "outputs": [],
   "source": [
    "groups = df[\"Country\"].to_numpy()\n",
    "X = df.drop([\"condition\",\"Unnamed: 0\", \"Country\"], axis=1)\n",
    "y = df[\"condition\"].to_numpy()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "1fe4d4bc",
   "metadata": {
    "scrolled": false
   },
   "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>Streptococcus_anginosus___ref_mOTU_v2_0004__</th>\n",
       "      <th>Enterobacteriaceae_sp.___ref_mOTU_v2_0036__</th>\n",
       "      <th>Citrobacter_sp.___ref_mOTU_v2_0076__</th>\n",
       "      <th>Klebsiella_michiganensis/oxytoca___ref_mOTU_v2_0079__</th>\n",
       "      <th>Enterococcus_faecalis___ref_mOTU_v2_0116__</th>\n",
       "      <th>Lactobacillus_salivarius___ref_mOTU_v2_0125__</th>\n",
       "      <th>Dielma_fastidiosa___ref_mOTU_v2_0138__</th>\n",
       "      <th>Streptococcus_constellatus/intermedius___ref_mOTU_v2_0143__</th>\n",
       "      <th>Streptococcus_parasanguinis___ref_mOTU_v2_0144__</th>\n",
       "      <th>Streptococcus_sp._HSISM1___ref_mOTU_v2_0145__</th>\n",
       "      <th>...</th>\n",
       "      <th>unknown_Porphyromonas___meta_mOTU_v2_7777__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7778__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7781__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7782__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7784__</th>\n",
       "      <th>Clostridium_sp._CAG__230___meta_mOTU_v2_7788__</th>\n",
       "      <th>Clostridium_sp._CAG__1193___meta_mOTU_v2_7789__</th>\n",
       "      <th>unknown_Erysipelotrichaceae___meta_mOTU_v2_7790__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7795__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7800__</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000083</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000249</td>\n",
       "      <td>0.000249</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000083</td>\n",
       "      <td>0.000083</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>0.000591</td>\n",
       "      <td>0.000066</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000066</td>\n",
       "      <td>0.002102</td>\n",
       "      <td>0.000197</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000788</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>0.000840</td>\n",
       "      <td>0.000065</td>\n",
       "      <td>0.002454</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000065</td>\n",
       "      <td>0.013111</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000065</td>\n",
       "      <td>0.012013</td>\n",
       "      <td>0.002777</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000194</td>\n",
       "      <td>0.000065</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000840</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.001808</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000071</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000213</td>\n",
       "      <td>0.000071</td>\n",
       "      <td>0.000142</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.000086</td>\n",
       "      <td>0.019236</td>\n",
       "      <td>0.001111</td>\n",
       "      <td>0.000256</td>\n",
       "      <td>0.000086</td>\n",
       "      <td>0.000086</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000513</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000171</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>763</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000118</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000059</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000118</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000059</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.002590</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.001118</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>764</th>\n",
       "      <td>0.000038</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000075</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>0.000113</td>\n",
       "      <td>0.000113</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000151</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000151</td>\n",
       "      <td>0.000038</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>765</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000348</td>\n",
       "      <td>0.000044</td>\n",
       "      <td>0.001741</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000435</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000044</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>766</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000312</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000208</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000208</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000156</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>767</th>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000784</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000087</td>\n",
       "      <td>0.000087</td>\n",
       "      <td>...</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.0</td>\n",
       "      <td>0.000000</td>\n",
       "      <td>0.000000</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>768 rows × 849 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Streptococcus_anginosus___ref_mOTU_v2_0004__  \\\n",
       "0                                        0.000000   \n",
       "1                                        0.000591   \n",
       "2                                        0.000840   \n",
       "3                                        0.000000   \n",
       "4                                        0.000086   \n",
       "..                                            ...   \n",
       "763                                      0.000000   \n",
       "764                                      0.000038   \n",
       "765                                      0.000000   \n",
       "766                                      0.000000   \n",
       "767                                      0.000000   \n",
       "\n",
       "     Enterobacteriaceae_sp.___ref_mOTU_v2_0036__  \\\n",
       "0                                       0.000000   \n",
       "1                                       0.000066   \n",
       "2                                       0.000065   \n",
       "3                                       0.000071   \n",
       "4                                       0.019236   \n",
       "..                                           ...   \n",
       "763                                     0.000118   \n",
       "764                                     0.000038   \n",
       "765                                     0.000000   \n",
       "766                                     0.000312   \n",
       "767                                     0.000000   \n",
       "\n",
       "     Citrobacter_sp.___ref_mOTU_v2_0076__  \\\n",
       "0                                0.000000   \n",
       "1                                0.000000   \n",
       "2                                0.002454   \n",
       "3                                0.000000   \n",
       "4                                0.001111   \n",
       "..                                    ...   \n",
       "763                              0.000000   \n",
       "764                              0.000000   \n",
       "765                              0.000000   \n",
       "766                              0.000000   \n",
       "767                              0.000784   \n",
       "\n",
       "     Klebsiella_michiganensis/oxytoca___ref_mOTU_v2_0079__  \\\n",
       "0                                             0.000000       \n",
       "1                                             0.000000       \n",
       "2                                             0.000000       \n",
       "3                                             0.000000       \n",
       "4                                             0.000256       \n",
       "..                                                 ...       \n",
       "763                                           0.000000       \n",
       "764                                           0.000000       \n",
       "765                                           0.000000       \n",
       "766                                           0.000000       \n",
       "767                                           0.000000       \n",
       "\n",
       "     Enterococcus_faecalis___ref_mOTU_v2_0116__  \\\n",
       "0                                      0.000000   \n",
       "1                                      0.000000   \n",
       "2                                      0.000065   \n",
       "3                                      0.000000   \n",
       "4                                      0.000086   \n",
       "..                                          ...   \n",
       "763                                    0.000059   \n",
       "764                                    0.000000   \n",
       "765                                    0.000000   \n",
       "766                                    0.000208   \n",
       "767                                    0.000000   \n",
       "\n",
       "     Lactobacillus_salivarius___ref_mOTU_v2_0125__  \\\n",
       "0                                         0.000000   \n",
       "1                                         0.000000   \n",
       "2                                         0.013111   \n",
       "3                                         0.000000   \n",
       "4                                         0.000086   \n",
       "..                                             ...   \n",
       "763                                       0.000000   \n",
       "764                                       0.000000   \n",
       "765                                       0.000000   \n",
       "766                                       0.000000   \n",
       "767                                       0.000000   \n",
       "\n",
       "     Dielma_fastidiosa___ref_mOTU_v2_0138__  \\\n",
       "0                                  0.000083   \n",
       "1                                  0.000000   \n",
       "2                                  0.000000   \n",
       "3                                  0.000213   \n",
       "4                                  0.000000   \n",
       "..                                      ...   \n",
       "763                                0.000118   \n",
       "764                                0.000075   \n",
       "765                                0.000000   \n",
       "766                                0.000000   \n",
       "767                                0.000000   \n",
       "\n",
       "     Streptococcus_constellatus/intermedius___ref_mOTU_v2_0143__  \\\n",
       "0                                             0.000000             \n",
       "1                                             0.000066             \n",
       "2                                             0.000065             \n",
       "3                                             0.000071             \n",
       "4                                             0.000000             \n",
       "..                                                 ...             \n",
       "763                                           0.000000             \n",
       "764                                           0.000038             \n",
       "765                                           0.000000             \n",
       "766                                           0.000000             \n",
       "767                                           0.000000             \n",
       "\n",
       "     Streptococcus_parasanguinis___ref_mOTU_v2_0144__  \\\n",
       "0                                            0.000249   \n",
       "1                                            0.002102   \n",
       "2                                            0.012013   \n",
       "3                                            0.000142   \n",
       "4                                            0.000513   \n",
       "..                                                ...   \n",
       "763                                          0.000059   \n",
       "764                                          0.000113   \n",
       "765                                          0.000000   \n",
       "766                                          0.000000   \n",
       "767                                          0.000087   \n",
       "\n",
       "     Streptococcus_sp._HSISM1___ref_mOTU_v2_0145__  ...  \\\n",
       "0                                         0.000249  ...   \n",
       "1                                         0.000197  ...   \n",
       "2                                         0.002777  ...   \n",
       "3                                         0.000000  ...   \n",
       "4                                         0.000000  ...   \n",
       "..                                             ...  ...   \n",
       "763                                       0.000000  ...   \n",
       "764                                       0.000113  ...   \n",
       "765                                       0.000000  ...   \n",
       "766                                       0.000000  ...   \n",
       "767                                       0.000087  ...   \n",
       "\n",
       "     unknown_Porphyromonas___meta_mOTU_v2_7777__  \\\n",
       "0                                       0.000000   \n",
       "1                                       0.000000   \n",
       "2                                       0.000000   \n",
       "3                                       0.000000   \n",
       "4                                       0.000000   \n",
       "..                                           ...   \n",
       "763                                     0.002590   \n",
       "764                                     0.000151   \n",
       "765                                     0.000000   \n",
       "766                                     0.000000   \n",
       "767                                     0.000000   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7778__  \\\n",
       "0                                       0.000000   \n",
       "1                                       0.000000   \n",
       "2                                       0.000000   \n",
       "3                                       0.000000   \n",
       "4                                       0.000171   \n",
       "..                                           ...   \n",
       "763                                     0.000000   \n",
       "764                                     0.000000   \n",
       "765                                     0.000348   \n",
       "766                                     0.000000   \n",
       "767                                     0.000000   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7781__  \\\n",
       "0                                       0.000000   \n",
       "1                                       0.000000   \n",
       "2                                       0.000194   \n",
       "3                                       0.000000   \n",
       "4                                       0.000000   \n",
       "..                                           ...   \n",
       "763                                     0.000000   \n",
       "764                                     0.000000   \n",
       "765                                     0.000044   \n",
       "766                                     0.000000   \n",
       "767                                     0.000000   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7782__  \\\n",
       "0                                       0.000000   \n",
       "1                                       0.000788   \n",
       "2                                       0.000065   \n",
       "3                                       0.000000   \n",
       "4                                       0.000000   \n",
       "..                                           ...   \n",
       "763                                     0.001118   \n",
       "764                                     0.000151   \n",
       "765                                     0.001741   \n",
       "766                                     0.000000   \n",
       "767                                     0.000000   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7784__  \\\n",
       "0                                       0.000000   \n",
       "1                                       0.000000   \n",
       "2                                       0.000000   \n",
       "3                                       0.000000   \n",
       "4                                       0.000000   \n",
       "..                                           ...   \n",
       "763                                     0.000000   \n",
       "764                                     0.000038   \n",
       "765                                     0.000000   \n",
       "766                                     0.000208   \n",
       "767                                     0.000000   \n",
       "\n",
       "     Clostridium_sp._CAG__230___meta_mOTU_v2_7788__  \\\n",
       "0                                          0.000000   \n",
       "1                                          0.000000   \n",
       "2                                          0.000840   \n",
       "3                                          0.000000   \n",
       "4                                          0.000000   \n",
       "..                                              ...   \n",
       "763                                        0.000000   \n",
       "764                                        0.000000   \n",
       "765                                        0.000435   \n",
       "766                                        0.000000   \n",
       "767                                        0.000000   \n",
       "\n",
       "     Clostridium_sp._CAG__1193___meta_mOTU_v2_7789__  \\\n",
       "0                                                0.0   \n",
       "1                                                0.0   \n",
       "2                                                0.0   \n",
       "3                                                0.0   \n",
       "4                                                0.0   \n",
       "..                                               ...   \n",
       "763                                              0.0   \n",
       "764                                              0.0   \n",
       "765                                              0.0   \n",
       "766                                              0.0   \n",
       "767                                              0.0   \n",
       "\n",
       "     unknown_Erysipelotrichaceae___meta_mOTU_v2_7790__  \\\n",
       "0                                                  0.0   \n",
       "1                                                  0.0   \n",
       "2                                                  0.0   \n",
       "3                                                  0.0   \n",
       "4                                                  0.0   \n",
       "..                                                 ...   \n",
       "763                                                0.0   \n",
       "764                                                0.0   \n",
       "765                                                0.0   \n",
       "766                                                0.0   \n",
       "767                                                0.0   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7795__  \\\n",
       "0                                       0.000083   \n",
       "1                                       0.000000   \n",
       "2                                       0.001808   \n",
       "3                                       0.000000   \n",
       "4                                       0.000000   \n",
       "..                                           ...   \n",
       "763                                     0.000000   \n",
       "764                                     0.000000   \n",
       "765                                     0.000000   \n",
       "766                                     0.000000   \n",
       "767                                     0.000000   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7800__  \n",
       "0                                       0.000083  \n",
       "1                                       0.000000  \n",
       "2                                       0.000000  \n",
       "3                                       0.000000  \n",
       "4                                       0.000000  \n",
       "..                                           ...  \n",
       "763                                     0.000000  \n",
       "764                                     0.000000  \n",
       "765                                     0.000044  \n",
       "766                                     0.000156  \n",
       "767                                     0.000000  \n",
       "\n",
       "[768 rows x 849 columns]"
      ]
     },
     "execution_count": 11,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "42222761",
   "metadata": {},
   "source": [
    "# Pandas dataframe to R table"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "088723d8",
   "metadata": {},
   "outputs": [],
   "source": [
    "with localconverter(robjects.default_converter + pandas2ri.converter):\n",
    "    r_from_pd_df = robjects.conversion.py2rpy(X)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "0244b3ab",
   "metadata": {},
   "source": [
    "# Zero replacement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "2096b00c",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "No. adjusted imputations:  425572 \n"
     ]
    }
   ],
   "source": [
    "np.random.seed(333)\n",
    "R_X = zCompositions.cmultRepl(r_from_pd_df, label = 0)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "5e3d73ca",
   "metadata": {},
   "source": [
    "# CLR"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "da946733",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(333)\n",
    "R_clr_X = robCompositions.cenLR(R_X)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "1cc67ebe",
   "metadata": {},
   "source": [
    "# R table to Pandas dataframe"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "7fd8bdc0",
   "metadata": {},
   "outputs": [],
   "source": [
    "with localconverter(robjects.default_converter + pandas2ri.converter):\n",
    "    X = robjects.conversion.rpy2py(R_clr_X[0])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "id": "c73fedcd",
   "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>Streptococcus_anginosus___ref_mOTU_v2_0004__</th>\n",
       "      <th>Enterobacteriaceae_sp.___ref_mOTU_v2_0036__</th>\n",
       "      <th>Citrobacter_sp.___ref_mOTU_v2_0076__</th>\n",
       "      <th>Klebsiella_michiganensis/oxytoca___ref_mOTU_v2_0079__</th>\n",
       "      <th>Enterococcus_faecalis___ref_mOTU_v2_0116__</th>\n",
       "      <th>Lactobacillus_salivarius___ref_mOTU_v2_0125__</th>\n",
       "      <th>Dielma_fastidiosa___ref_mOTU_v2_0138__</th>\n",
       "      <th>Streptococcus_constellatus/intermedius___ref_mOTU_v2_0143__</th>\n",
       "      <th>Streptococcus_parasanguinis___ref_mOTU_v2_0144__</th>\n",
       "      <th>Streptococcus_sp._HSISM1___ref_mOTU_v2_0145__</th>\n",
       "      <th>...</th>\n",
       "      <th>unknown_Porphyromonas___meta_mOTU_v2_7777__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7778__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7781__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7782__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7784__</th>\n",
       "      <th>Clostridium_sp._CAG__230___meta_mOTU_v2_7788__</th>\n",
       "      <th>Clostridium_sp._CAG__1193___meta_mOTU_v2_7789__</th>\n",
       "      <th>unknown_Erysipelotrichaceae___meta_mOTU_v2_7790__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7795__</th>\n",
       "      <th>unknown_Clostridiales___meta_mOTU_v2_7800__</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>0</th>\n",
       "      <td>-0.639814</td>\n",
       "      <td>-1.123042</td>\n",
       "      <td>-0.638052</td>\n",
       "      <td>-0.627669</td>\n",
       "      <td>-0.766950</td>\n",
       "      <td>-0.451501</td>\n",
       "      <td>0.657708</td>\n",
       "      <td>-0.866839</td>\n",
       "      <td>1.756421</td>\n",
       "      <td>1.756421</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.572469</td>\n",
       "      <td>-0.712546</td>\n",
       "      <td>-0.536782</td>\n",
       "      <td>-0.724254</td>\n",
       "      <td>-0.560843</td>\n",
       "      <td>-0.535670</td>\n",
       "      <td>-0.353747</td>\n",
       "      <td>-0.392983</td>\n",
       "      <td>0.657708</td>\n",
       "      <td>0.657708</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1</th>\n",
       "      <td>2.453491</td>\n",
       "      <td>0.256614</td>\n",
       "      <td>-0.833799</td>\n",
       "      <td>-0.823416</td>\n",
       "      <td>-0.962697</td>\n",
       "      <td>-0.647248</td>\n",
       "      <td>-0.873193</td>\n",
       "      <td>0.256614</td>\n",
       "      <td>3.722002</td>\n",
       "      <td>1.354877</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.768216</td>\n",
       "      <td>-0.908293</td>\n",
       "      <td>-0.732529</td>\n",
       "      <td>2.741174</td>\n",
       "      <td>-0.756590</td>\n",
       "      <td>-0.731417</td>\n",
       "      <td>-0.549494</td>\n",
       "      <td>-0.588730</td>\n",
       "      <td>-0.970122</td>\n",
       "      <td>-0.962697</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>2</th>\n",
       "      <td>2.598610</td>\n",
       "      <td>0.033862</td>\n",
       "      <td>3.671247</td>\n",
       "      <td>-1.061800</td>\n",
       "      <td>0.033862</td>\n",
       "      <td>5.346866</td>\n",
       "      <td>-1.111577</td>\n",
       "      <td>0.033862</td>\n",
       "      <td>5.259407</td>\n",
       "      <td>3.794861</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.006600</td>\n",
       "      <td>-1.146677</td>\n",
       "      <td>1.132273</td>\n",
       "      <td>0.033862</td>\n",
       "      <td>-0.994974</td>\n",
       "      <td>2.598610</td>\n",
       "      <td>-0.787878</td>\n",
       "      <td>-0.827114</td>\n",
       "      <td>3.365865</td>\n",
       "      <td>-1.201081</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>3</th>\n",
       "      <td>-0.596255</td>\n",
       "      <td>0.544960</td>\n",
       "      <td>-0.594493</td>\n",
       "      <td>-0.584110</td>\n",
       "      <td>-0.723391</td>\n",
       "      <td>-0.407942</td>\n",
       "      <td>1.643102</td>\n",
       "      <td>0.544960</td>\n",
       "      <td>1.237634</td>\n",
       "      <td>-0.723391</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.528910</td>\n",
       "      <td>-0.668988</td>\n",
       "      <td>-0.493223</td>\n",
       "      <td>-0.680695</td>\n",
       "      <td>-0.517284</td>\n",
       "      <td>-0.492111</td>\n",
       "      <td>-0.310188</td>\n",
       "      <td>-0.349424</td>\n",
       "      <td>-0.730816</td>\n",
       "      <td>-0.723391</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>4</th>\n",
       "      <td>0.157178</td>\n",
       "      <td>5.573185</td>\n",
       "      <td>2.722033</td>\n",
       "      <td>1.255696</td>\n",
       "      <td>0.157178</td>\n",
       "      <td>0.157178</td>\n",
       "      <td>-1.229870</td>\n",
       "      <td>-1.419263</td>\n",
       "      <td>1.948844</td>\n",
       "      <td>-1.319373</td>\n",
       "      <td>...</td>\n",
       "      <td>-1.124892</td>\n",
       "      <td>0.850231</td>\n",
       "      <td>-1.089205</td>\n",
       "      <td>-1.276677</td>\n",
       "      <td>-1.113267</td>\n",
       "      <td>-1.088093</td>\n",
       "      <td>-0.906170</td>\n",
       "      <td>-0.945406</td>\n",
       "      <td>-1.326798</td>\n",
       "      <td>-1.319373</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>763</th>\n",
       "      <td>-0.996471</td>\n",
       "      <td>0.650929</td>\n",
       "      <td>-0.994709</td>\n",
       "      <td>-0.984326</td>\n",
       "      <td>-0.041564</td>\n",
       "      <td>-0.808158</td>\n",
       "      <td>0.650929</td>\n",
       "      <td>-1.223496</td>\n",
       "      <td>-0.041564</td>\n",
       "      <td>-1.123607</td>\n",
       "      <td>...</td>\n",
       "      <td>3.741974</td>\n",
       "      <td>-1.069203</td>\n",
       "      <td>-0.893439</td>\n",
       "      <td>2.902223</td>\n",
       "      <td>-0.917500</td>\n",
       "      <td>-0.892327</td>\n",
       "      <td>-0.710404</td>\n",
       "      <td>-0.749640</td>\n",
       "      <td>-1.131032</td>\n",
       "      <td>-1.123607</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>764</th>\n",
       "      <td>-0.490840</td>\n",
       "      <td>-0.490840</td>\n",
       "      <td>-1.000097</td>\n",
       "      <td>-0.989714</td>\n",
       "      <td>-1.128995</td>\n",
       "      <td>-0.813546</td>\n",
       "      <td>0.203636</td>\n",
       "      <td>-0.490840</td>\n",
       "      <td>0.608517</td>\n",
       "      <td>0.608517</td>\n",
       "      <td>...</td>\n",
       "      <td>0.896199</td>\n",
       "      <td>-1.074592</td>\n",
       "      <td>-0.898827</td>\n",
       "      <td>0.896199</td>\n",
       "      <td>-0.490840</td>\n",
       "      <td>-0.897715</td>\n",
       "      <td>-0.715792</td>\n",
       "      <td>-0.755028</td>\n",
       "      <td>-1.136420</td>\n",
       "      <td>-1.128995</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>765</th>\n",
       "      <td>-0.996933</td>\n",
       "      <td>-1.480161</td>\n",
       "      <td>-0.995171</td>\n",
       "      <td>-0.984788</td>\n",
       "      <td>-1.124069</td>\n",
       "      <td>-0.808619</td>\n",
       "      <td>-1.034565</td>\n",
       "      <td>-1.223958</td>\n",
       "      <td>-1.404673</td>\n",
       "      <td>-1.124069</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.929588</td>\n",
       "      <td>1.756708</td>\n",
       "      <td>-0.323276</td>\n",
       "      <td>3.366147</td>\n",
       "      <td>-0.917962</td>\n",
       "      <td>1.979854</td>\n",
       "      <td>-0.710866</td>\n",
       "      <td>-0.750102</td>\n",
       "      <td>-1.131494</td>\n",
       "      <td>-0.323276</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>766</th>\n",
       "      <td>-0.751492</td>\n",
       "      <td>1.998640</td>\n",
       "      <td>-0.749730</td>\n",
       "      <td>-0.739347</td>\n",
       "      <td>1.593176</td>\n",
       "      <td>-0.563179</td>\n",
       "      <td>-0.789125</td>\n",
       "      <td>-0.978518</td>\n",
       "      <td>-1.159233</td>\n",
       "      <td>-0.878628</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.684147</td>\n",
       "      <td>-0.824225</td>\n",
       "      <td>-0.648460</td>\n",
       "      <td>-0.835932</td>\n",
       "      <td>1.593176</td>\n",
       "      <td>-0.647348</td>\n",
       "      <td>-0.465425</td>\n",
       "      <td>-0.504661</td>\n",
       "      <td>-0.886053</td>\n",
       "      <td>1.305489</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>767</th>\n",
       "      <td>-0.619330</td>\n",
       "      <td>-1.102559</td>\n",
       "      <td>2.941765</td>\n",
       "      <td>-0.607185</td>\n",
       "      <td>-0.746466</td>\n",
       "      <td>-0.431017</td>\n",
       "      <td>-0.656963</td>\n",
       "      <td>-0.846356</td>\n",
       "      <td>0.744883</td>\n",
       "      <td>0.744883</td>\n",
       "      <td>...</td>\n",
       "      <td>-0.551985</td>\n",
       "      <td>-0.692063</td>\n",
       "      <td>-0.516298</td>\n",
       "      <td>-0.703770</td>\n",
       "      <td>-0.540359</td>\n",
       "      <td>-0.515186</td>\n",
       "      <td>-0.333263</td>\n",
       "      <td>-0.372499</td>\n",
       "      <td>-0.753891</td>\n",
       "      <td>-0.746466</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>768 rows × 849 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "     Streptococcus_anginosus___ref_mOTU_v2_0004__  \\\n",
       "0                                       -0.639814   \n",
       "1                                        2.453491   \n",
       "2                                        2.598610   \n",
       "3                                       -0.596255   \n",
       "4                                        0.157178   \n",
       "..                                            ...   \n",
       "763                                     -0.996471   \n",
       "764                                     -0.490840   \n",
       "765                                     -0.996933   \n",
       "766                                     -0.751492   \n",
       "767                                     -0.619330   \n",
       "\n",
       "     Enterobacteriaceae_sp.___ref_mOTU_v2_0036__  \\\n",
       "0                                      -1.123042   \n",
       "1                                       0.256614   \n",
       "2                                       0.033862   \n",
       "3                                       0.544960   \n",
       "4                                       5.573185   \n",
       "..                                           ...   \n",
       "763                                     0.650929   \n",
       "764                                    -0.490840   \n",
       "765                                    -1.480161   \n",
       "766                                     1.998640   \n",
       "767                                    -1.102559   \n",
       "\n",
       "     Citrobacter_sp.___ref_mOTU_v2_0076__  \\\n",
       "0                               -0.638052   \n",
       "1                               -0.833799   \n",
       "2                                3.671247   \n",
       "3                               -0.594493   \n",
       "4                                2.722033   \n",
       "..                                    ...   \n",
       "763                             -0.994709   \n",
       "764                             -1.000097   \n",
       "765                             -0.995171   \n",
       "766                             -0.749730   \n",
       "767                              2.941765   \n",
       "\n",
       "     Klebsiella_michiganensis/oxytoca___ref_mOTU_v2_0079__  \\\n",
       "0                                            -0.627669       \n",
       "1                                            -0.823416       \n",
       "2                                            -1.061800       \n",
       "3                                            -0.584110       \n",
       "4                                             1.255696       \n",
       "..                                                 ...       \n",
       "763                                          -0.984326       \n",
       "764                                          -0.989714       \n",
       "765                                          -0.984788       \n",
       "766                                          -0.739347       \n",
       "767                                          -0.607185       \n",
       "\n",
       "     Enterococcus_faecalis___ref_mOTU_v2_0116__  \\\n",
       "0                                     -0.766950   \n",
       "1                                     -0.962697   \n",
       "2                                      0.033862   \n",
       "3                                     -0.723391   \n",
       "4                                      0.157178   \n",
       "..                                          ...   \n",
       "763                                   -0.041564   \n",
       "764                                   -1.128995   \n",
       "765                                   -1.124069   \n",
       "766                                    1.593176   \n",
       "767                                   -0.746466   \n",
       "\n",
       "     Lactobacillus_salivarius___ref_mOTU_v2_0125__  \\\n",
       "0                                        -0.451501   \n",
       "1                                        -0.647248   \n",
       "2                                         5.346866   \n",
       "3                                        -0.407942   \n",
       "4                                         0.157178   \n",
       "..                                             ...   \n",
       "763                                      -0.808158   \n",
       "764                                      -0.813546   \n",
       "765                                      -0.808619   \n",
       "766                                      -0.563179   \n",
       "767                                      -0.431017   \n",
       "\n",
       "     Dielma_fastidiosa___ref_mOTU_v2_0138__  \\\n",
       "0                                  0.657708   \n",
       "1                                 -0.873193   \n",
       "2                                 -1.111577   \n",
       "3                                  1.643102   \n",
       "4                                 -1.229870   \n",
       "..                                      ...   \n",
       "763                                0.650929   \n",
       "764                                0.203636   \n",
       "765                               -1.034565   \n",
       "766                               -0.789125   \n",
       "767                               -0.656963   \n",
       "\n",
       "     Streptococcus_constellatus/intermedius___ref_mOTU_v2_0143__  \\\n",
       "0                                            -0.866839             \n",
       "1                                             0.256614             \n",
       "2                                             0.033862             \n",
       "3                                             0.544960             \n",
       "4                                            -1.419263             \n",
       "..                                                 ...             \n",
       "763                                          -1.223496             \n",
       "764                                          -0.490840             \n",
       "765                                          -1.223958             \n",
       "766                                          -0.978518             \n",
       "767                                          -0.846356             \n",
       "\n",
       "     Streptococcus_parasanguinis___ref_mOTU_v2_0144__  \\\n",
       "0                                            1.756421   \n",
       "1                                            3.722002   \n",
       "2                                            5.259407   \n",
       "3                                            1.237634   \n",
       "4                                            1.948844   \n",
       "..                                                ...   \n",
       "763                                         -0.041564   \n",
       "764                                          0.608517   \n",
       "765                                         -1.404673   \n",
       "766                                         -1.159233   \n",
       "767                                          0.744883   \n",
       "\n",
       "     Streptococcus_sp._HSISM1___ref_mOTU_v2_0145__  ...  \\\n",
       "0                                         1.756421  ...   \n",
       "1                                         1.354877  ...   \n",
       "2                                         3.794861  ...   \n",
       "3                                        -0.723391  ...   \n",
       "4                                        -1.319373  ...   \n",
       "..                                             ...  ...   \n",
       "763                                      -1.123607  ...   \n",
       "764                                       0.608517  ...   \n",
       "765                                      -1.124069  ...   \n",
       "766                                      -0.878628  ...   \n",
       "767                                       0.744883  ...   \n",
       "\n",
       "     unknown_Porphyromonas___meta_mOTU_v2_7777__  \\\n",
       "0                                      -0.572469   \n",
       "1                                      -0.768216   \n",
       "2                                      -1.006600   \n",
       "3                                      -0.528910   \n",
       "4                                      -1.124892   \n",
       "..                                           ...   \n",
       "763                                     3.741974   \n",
       "764                                     0.896199   \n",
       "765                                    -0.929588   \n",
       "766                                    -0.684147   \n",
       "767                                    -0.551985   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7778__  \\\n",
       "0                                      -0.712546   \n",
       "1                                      -0.908293   \n",
       "2                                      -1.146677   \n",
       "3                                      -0.668988   \n",
       "4                                       0.850231   \n",
       "..                                           ...   \n",
       "763                                    -1.069203   \n",
       "764                                    -1.074592   \n",
       "765                                     1.756708   \n",
       "766                                    -0.824225   \n",
       "767                                    -0.692063   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7781__  \\\n",
       "0                                      -0.536782   \n",
       "1                                      -0.732529   \n",
       "2                                       1.132273   \n",
       "3                                      -0.493223   \n",
       "4                                      -1.089205   \n",
       "..                                           ...   \n",
       "763                                    -0.893439   \n",
       "764                                    -0.898827   \n",
       "765                                    -0.323276   \n",
       "766                                    -0.648460   \n",
       "767                                    -0.516298   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7782__  \\\n",
       "0                                      -0.724254   \n",
       "1                                       2.741174   \n",
       "2                                       0.033862   \n",
       "3                                      -0.680695   \n",
       "4                                      -1.276677   \n",
       "..                                           ...   \n",
       "763                                     2.902223   \n",
       "764                                     0.896199   \n",
       "765                                     3.366147   \n",
       "766                                    -0.835932   \n",
       "767                                    -0.703770   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7784__  \\\n",
       "0                                      -0.560843   \n",
       "1                                      -0.756590   \n",
       "2                                      -0.994974   \n",
       "3                                      -0.517284   \n",
       "4                                      -1.113267   \n",
       "..                                           ...   \n",
       "763                                    -0.917500   \n",
       "764                                    -0.490840   \n",
       "765                                    -0.917962   \n",
       "766                                     1.593176   \n",
       "767                                    -0.540359   \n",
       "\n",
       "     Clostridium_sp._CAG__230___meta_mOTU_v2_7788__  \\\n",
       "0                                         -0.535670   \n",
       "1                                         -0.731417   \n",
       "2                                          2.598610   \n",
       "3                                         -0.492111   \n",
       "4                                         -1.088093   \n",
       "..                                              ...   \n",
       "763                                       -0.892327   \n",
       "764                                       -0.897715   \n",
       "765                                        1.979854   \n",
       "766                                       -0.647348   \n",
       "767                                       -0.515186   \n",
       "\n",
       "     Clostridium_sp._CAG__1193___meta_mOTU_v2_7789__  \\\n",
       "0                                          -0.353747   \n",
       "1                                          -0.549494   \n",
       "2                                          -0.787878   \n",
       "3                                          -0.310188   \n",
       "4                                          -0.906170   \n",
       "..                                               ...   \n",
       "763                                        -0.710404   \n",
       "764                                        -0.715792   \n",
       "765                                        -0.710866   \n",
       "766                                        -0.465425   \n",
       "767                                        -0.333263   \n",
       "\n",
       "     unknown_Erysipelotrichaceae___meta_mOTU_v2_7790__  \\\n",
       "0                                            -0.392983   \n",
       "1                                            -0.588730   \n",
       "2                                            -0.827114   \n",
       "3                                            -0.349424   \n",
       "4                                            -0.945406   \n",
       "..                                                 ...   \n",
       "763                                          -0.749640   \n",
       "764                                          -0.755028   \n",
       "765                                          -0.750102   \n",
       "766                                          -0.504661   \n",
       "767                                          -0.372499   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7795__  \\\n",
       "0                                       0.657708   \n",
       "1                                      -0.970122   \n",
       "2                                       3.365865   \n",
       "3                                      -0.730816   \n",
       "4                                      -1.326798   \n",
       "..                                           ...   \n",
       "763                                    -1.131032   \n",
       "764                                    -1.136420   \n",
       "765                                    -1.131494   \n",
       "766                                    -0.886053   \n",
       "767                                    -0.753891   \n",
       "\n",
       "     unknown_Clostridiales___meta_mOTU_v2_7800__  \n",
       "0                                       0.657708  \n",
       "1                                      -0.962697  \n",
       "2                                      -1.201081  \n",
       "3                                      -0.723391  \n",
       "4                                      -1.319373  \n",
       "..                                           ...  \n",
       "763                                    -1.123607  \n",
       "764                                    -1.128995  \n",
       "765                                    -0.323276  \n",
       "766                                     1.305489  \n",
       "767                                    -0.746466  \n",
       "\n",
       "[768 rows x 849 columns]"
      ]
     },
     "execution_count": 16,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "X"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "874330dc",
   "metadata": {},
   "source": [
    "## Dataset names"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "id": "3d31b47b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "['AT-CRC_species',\n",
       " 'CN-CRC_species',\n",
       " 'DE-CRC_species',\n",
       " 'FR-CRC_species',\n",
       " 'IT-CRC-2_species',\n",
       " 'IT-CRC_species',\n",
       " 'JP-CRC_species',\n",
       " 'US-CRC_species']"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "unique_groups = list(np.unique(groups))\n",
    "unique_groups"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "id": "be80a857",
   "metadata": {
    "scrolled": true
   },
   "outputs": [],
   "source": [
    "AUC_scores = pd.DataFrame(index=unique_groups, columns=unique_groups)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "id": "21e98d1d",
   "metadata": {},
   "outputs": [],
   "source": [
    "ROC_curve = pd.DataFrame(index=unique_groups, columns=[\"fpr\",\"tpr\",\"auc\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "983da394",
   "metadata": {},
   "source": [
    "# LOSO validation with random forests"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "id": "149e8155",
   "metadata": {},
   "outputs": [],
   "source": [
    "SEED = 123\n",
    "show_fold_stats = True"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "id": "44d0a12c",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--------------------------------------------------------------------------------\n",
      "Fold : 1 CN-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8946\n",
      "F1 : 0.7826, Recall : 0.8333 , Precision : 0.7377\n",
      "[[58 16]\n",
      " [ 9 45]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 2 DE-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8161\n",
      "F1 : 0.7692, Recall : 0.8333 , Precision : 0.7143\n",
      "[[40 20]\n",
      " [10 50]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 3 FR-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8107\n",
      "F1 : 0.7103, Recall : 0.6230 , Precision : 0.8261\n",
      "[[45  8]\n",
      " [23 38]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 4 IT-CRC-2_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7388\n",
      "F1 : 0.6552, Recall : 0.6786 , Precision : 0.6333\n",
      "[[21 11]\n",
      " [ 9 19]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 5 IT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7823\n",
      "F1 : 0.6364, Recall : 0.5833 , Precision : 0.7000\n",
      "[[23  6]\n",
      " [10 14]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 6 JP-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8300\n",
      "F1 : 0.6250, Recall : 0.5000 , Precision : 0.8333\n",
      "[[36  4]\n",
      " [20 20]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 7 US-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7539\n",
      "F1 : 0.6471, Recall : 0.6346 , Precision : 0.6600\n",
      "[[35 17]\n",
      " [19 33]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Group : 1 AT-CRC_species\n",
      "ROC AUC score for RandomForest model, LOGO test set: 0.8687\n",
      "F1 : 0.8421, Recall : 0.8889 , Precision : 0.8000\n",
      "[[32 14]\n",
      " [ 7 56]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 1 AT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8742\n",
      "F1 : 0.8444, Recall : 0.9048 , Precision : 0.7917\n",
      "[[31 15]\n",
      " [ 6 57]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 2 DE-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7868\n",
      "F1 : 0.7258, Recall : 0.7500 , Precision : 0.7031\n",
      "[[41 19]\n",
      " [15 45]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 3 FR-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8174\n",
      "F1 : 0.7273, Recall : 0.7213 , Precision : 0.7333\n",
      "[[37 16]\n",
      " [17 44]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 4 IT-CRC-2_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7595\n",
      "F1 : 0.6984, Recall : 0.7857 , Precision : 0.6286\n",
      "[[19 13]\n",
      " [ 6 22]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 5 IT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7141\n",
      "F1 : 0.6250, Recall : 0.6250 , Precision : 0.6250\n",
      "[[20  9]\n",
      " [ 9 15]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 6 JP-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8566\n",
      "F1 : 0.7632, Recall : 0.7250 , Precision : 0.8056\n",
      "[[33  7]\n",
      " [11 29]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 7 US-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7274\n",
      "F1 : 0.7091, Recall : 0.7500 , Precision : 0.6724\n",
      "[[33 19]\n",
      " [13 39]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Group : 2 CN-CRC_species\n",
      "ROC AUC score for RandomForest model, LOGO test set: 0.9015\n",
      "F1 : 0.8065, Recall : 0.9259 , Precision : 0.7143\n",
      "[[54 20]\n",
      " [ 4 50]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 1 AT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8989\n",
      "F1 : 0.8358, Recall : 0.8889 , Precision : 0.7887\n",
      "[[31 15]\n",
      " [ 7 56]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 2 CN-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8780\n",
      "F1 : 0.7967, Recall : 0.9074 , Precision : 0.7101\n",
      "[[54 20]\n",
      " [ 5 49]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 3 FR-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8172\n",
      "F1 : 0.7458, Recall : 0.7213 , Precision : 0.7719\n",
      "[[40 13]\n",
      " [17 44]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 4 IT-CRC-2_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7723\n",
      "F1 : 0.6780, Recall : 0.7143 , Precision : 0.6452\n",
      "[[21 11]\n",
      " [ 8 20]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 5 IT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8118\n",
      "F1 : 0.5909, Recall : 0.5417 , Precision : 0.6500\n",
      "[[22  7]\n",
      " [11 13]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 6 JP-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8772\n",
      "F1 : 0.7778, Recall : 0.7000 , Precision : 0.8750\n",
      "[[36  4]\n",
      " [12 28]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 7 US-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.6762\n",
      "F1 : 0.5957, Recall : 0.5385 , Precision : 0.6667\n",
      "[[38 14]\n",
      " [24 28]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Group : 3 DE-CRC_species\n",
      "ROC AUC score for RandomForest model, LOGO test set: 0.8140\n",
      "F1 : 0.7742, Recall : 0.8000 , Precision : 0.7500\n",
      "[[44 16]\n",
      " [12 48]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 1 AT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8585\n",
      "F1 : 0.8000, Recall : 0.7619 , Precision : 0.8421\n",
      "[[37  9]\n",
      " [15 48]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 2 CN-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8860\n",
      "F1 : 0.8205, Recall : 0.8889 , Precision : 0.7619\n",
      "[[59 15]\n",
      " [ 6 48]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 3 DE-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8118\n",
      "F1 : 0.7939, Recall : 0.8667 , Precision : 0.7324\n",
      "[[41 19]\n",
      " [ 8 52]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 4 IT-CRC-2_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7885\n",
      "F1 : 0.7018, Recall : 0.7143 , Precision : 0.6897\n",
      "[[23  9]\n",
      " [ 8 20]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 5 IT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8283\n",
      "F1 : 0.7451, Recall : 0.7917 , Precision : 0.7037\n",
      "[[21  8]\n",
      " [ 5 19]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 6 JP-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8612\n",
      "F1 : 0.7324, Recall : 0.6500 , Precision : 0.8387\n",
      "[[35  5]\n",
      " [14 26]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 7 US-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7254\n",
      "F1 : 0.6535, Recall : 0.6346 , Precision : 0.6735\n",
      "[[36 16]\n",
      " [19 33]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Group : 4 FR-CRC_species\n",
      "ROC AUC score for RandomForest model, LOGO test set: 0.8181\n",
      "F1 : 0.7304, Recall : 0.6885 , Precision : 0.7778\n",
      "[[41 12]\n",
      " [19 42]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "--------------------------------------------------------------------------------\n",
      "Fold : 1 AT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8644\n",
      "F1 : 0.8613, Recall : 0.9365 , Precision : 0.7973\n",
      "[[31 15]\n",
      " [ 4 59]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 2 CN-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8943\n",
      "F1 : 0.7934, Recall : 0.8889 , Precision : 0.7164\n",
      "[[55 19]\n",
      " [ 6 48]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 3 DE-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8242\n",
      "F1 : 0.7752, Recall : 0.8333 , Precision : 0.7246\n",
      "[[41 19]\n",
      " [10 50]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 4 FR-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8067\n",
      "F1 : 0.7500, Recall : 0.7377 , Precision : 0.7627\n",
      "[[39 14]\n",
      " [16 45]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 5 IT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7780\n",
      "F1 : 0.6047, Recall : 0.5417 , Precision : 0.6842\n",
      "[[23  6]\n",
      " [11 13]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 6 JP-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8403\n",
      "F1 : 0.7324, Recall : 0.6500 , Precision : 0.8387\n",
      "[[35  5]\n",
      " [14 26]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 7 US-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7269\n",
      "F1 : 0.6667, Recall : 0.6538 , Precision : 0.6800\n",
      "[[36 16]\n",
      " [18 34]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Group : 5 IT-CRC-2_species\n",
      "ROC AUC score for RandomForest model, LOGO test set: 0.7584\n",
      "F1 : 0.7097, Recall : 0.7857 , Precision : 0.6471\n",
      "[[20 12]\n",
      " [ 6 22]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 1 AT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8637\n",
      "F1 : 0.8333, Recall : 0.8730 , Precision : 0.7971\n",
      "[[32 14]\n",
      " [ 8 55]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 2 CN-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8768\n",
      "F1 : 0.7840, Recall : 0.9074 , Precision : 0.6901\n",
      "[[52 22]\n",
      " [ 5 49]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 3 DE-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8285\n",
      "F1 : 0.7742, Recall : 0.8000 , Precision : 0.7500\n",
      "[[44 16]\n",
      " [12 48]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 4 FR-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8170\n",
      "F1 : 0.7521, Recall : 0.7213 , Precision : 0.7857\n",
      "[[41 12]\n",
      " [17 44]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 5 IT-CRC-2_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7740\n",
      "F1 : 0.6885, Recall : 0.7500 , Precision : 0.6364\n",
      "[[20 12]\n",
      " [ 7 21]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 6 JP-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8506\n",
      "F1 : 0.7536, Recall : 0.6500 , Precision : 0.8966\n",
      "[[37  3]\n",
      " [14 26]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 7 US-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7322\n",
      "F1 : 0.6275, Recall : 0.6154 , Precision : 0.6400\n",
      "[[34 18]\n",
      " [20 32]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Group : 6 IT-CRC_species\n",
      "ROC AUC score for RandomForest model, LOGO test set: 0.7888\n",
      "F1 : 0.6222, Recall : 0.5833 , Precision : 0.6667\n",
      "[[22  7]\n",
      " [10 14]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 1 AT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8582\n",
      "F1 : 0.8397, Recall : 0.8730 , Precision : 0.8088\n",
      "[[33 13]\n",
      " [ 8 55]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 2 CN-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8979\n",
      "F1 : 0.8034, Recall : 0.8704 , Precision : 0.7460\n",
      "[[58 16]\n",
      " [ 7 47]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 3 DE-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8214\n",
      "F1 : 0.7597, Recall : 0.8167 , Precision : 0.7101\n",
      "[[40 20]\n",
      " [11 49]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 4 FR-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8263\n",
      "F1 : 0.7368, Recall : 0.6885 , Precision : 0.7925\n",
      "[[42 11]\n",
      " [19 42]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 5 IT-CRC-2_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7612\n",
      "F1 : 0.6984, Recall : 0.7857 , Precision : 0.6286\n",
      "[[19 13]\n",
      " [ 6 22]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 6 IT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7723\n",
      "F1 : 0.5641, Recall : 0.4583 , Precision : 0.7333\n",
      "[[25  4]\n",
      " [13 11]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 7 US-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7398\n",
      "F1 : 0.6327, Recall : 0.5962 , Precision : 0.6739\n",
      "[[37 15]\n",
      " [21 31]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Group : 7 JP-CRC_species\n",
      "ROC AUC score for RandomForest model, LOGO test set: 0.8694\n",
      "F1 : 0.7647, Recall : 0.6500 , Precision : 0.9286\n",
      "[[38  2]\n",
      " [14 26]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 1 AT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8742\n",
      "F1 : 0.8333, Recall : 0.8730 , Precision : 0.7971\n",
      "[[32 14]\n",
      " [ 8 55]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 2 CN-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8739\n",
      "F1 : 0.7680, Recall : 0.8889 , Precision : 0.6761\n",
      "[[51 23]\n",
      " [ 6 48]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 3 DE-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7610\n",
      "F1 : 0.6496, Recall : 0.6333 , Precision : 0.6667\n",
      "[[41 19]\n",
      " [22 38]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 4 FR-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8223\n",
      "F1 : 0.7156, Recall : 0.6393 , Precision : 0.8125\n",
      "[[44  9]\n",
      " [22 39]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 5 IT-CRC-2_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.7673\n",
      "F1 : 0.6984, Recall : 0.7857 , Precision : 0.6286\n",
      "[[19 13]\n",
      " [ 6 22]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 6 IT-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8175\n",
      "F1 : 0.6512, Recall : 0.5833 , Precision : 0.7368\n",
      "[[24  5]\n",
      " [10 14]]\n",
      "--------------------------------------------------------------------------------\n",
      "Fold : 7 JP-CRC_species\n",
      "ROC AUC score for RandomForest model, validation set: 0.8559\n",
      "F1 : 0.8108, Recall : 0.7500 , Precision : 0.8824\n",
      "[[36  4]\n",
      " [10 30]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n",
      "Group : 8 US-CRC_species\n",
      "ROC AUC score for RandomForest model, LOGO test set: 0.7295\n",
      "F1 : 0.6731, Recall : 0.6731 , Precision : 0.6731\n",
      "[[35 17]\n",
      " [17 35]]\n",
      "--------------------------------------------------------------------------------\n",
      "--------------------------------------------------------------------------------\n"
     ]
    }
   ],
   "source": [
    "grp = 0\n",
    "for i in range(len(unique_groups)):\n",
    "    outer_test_index = list(np.where(df[\"Country\"]==unique_groups[i])[0])\n",
    "    \n",
    "    inner_group_list = []\n",
    "    inner_group_list.append(unique_groups[:i] + unique_groups[i+1:])\n",
    "    \n",
    "    outer_train_index = list(np.where(df[\"Country\"]!=unique_groups[i])[0])\n",
    "    \n",
    "    #Training dataset\n",
    "    outer_train_X = X.iloc[outer_train_index]\n",
    "    outer_train_y = y[outer_train_index]\n",
    "    \n",
    "    #Test dataset\n",
    "    outer_test_X =  X.iloc[outer_test_index]\n",
    "    outer_test_y =  y[outer_test_index]\n",
    "    \n",
    "    for k,name in enumerate(inner_group_list[0]):\n",
    "        inner_val_index = list(np.where(df[\"Country\"]==name)[0])\n",
    "        \n",
    "        inner_train_names = []\n",
    "        inner_train_names.append(inner_group_list[:k] + inner_group_list[k+1:])\n",
    "        \n",
    "        inner_train_index = list(np.where((df[\"Country\"]!=name) & (df[\"Country\"]!=unique_groups[i]))[0])\n",
    "    \n",
    "        #Inner training dataset\n",
    "        inner_train_X = X.iloc[inner_train_index]\n",
    "        inner_train_y = y[inner_train_index]\n",
    "        \n",
    "        #Inner validation dataset\n",
    "        inner_val_X =  X.iloc[inner_val_index]\n",
    "        inner_val_y =  y[inner_val_index]\n",
    "        \n",
    "               \n",
    "        # Random Forest\n",
    "        clf = RandomForestClassifier(n_estimators = 500, random_state = SEED, verbose = 0, bootstrap=True)\n",
    "        clf.fit(inner_train_X, inner_train_y)\n",
    "\n",
    "        RF_pred_class = clf.predict(inner_val_X)\n",
    "        RF_preds = clf.predict_proba(inner_val_X)\n",
    "\n",
    "        RF_AUC_test_score = roc_auc_score(inner_val_y, RF_preds[:,1])\n",
    "        RF_f1_test = f1_score(inner_val_y, RF_pred_class)\n",
    "        RF_recall_test = recall_score(inner_val_y, RF_pred_class)\n",
    "        RF_precision_test = precision_score(inner_val_y, RF_pred_class)\n",
    "        \n",
    "        #Specificity and sensitivity\n",
    "        tn, fp, fn, tp = confusion_matrix(inner_val_y, RF_pred_class).ravel()\n",
    "        specificity = tn / (tn + fp)\n",
    "        sensitivity = tp / (tp + fn)\n",
    "        \n",
    "        #Adding auc to dataframe\n",
    "        AUC_scores.at[unique_groups[i], name] = RF_AUC_test_score\n",
    "\n",
    "        if show_fold_stats:\n",
    "            print('-' * 80)\n",
    "            print('Fold : %s'%(k+1),name)\n",
    "            print('ROC AUC score for RandomForest model, validation set: %.4f'%RF_AUC_test_score)\n",
    "            print('F1 : %.4f, Recall : %.4f , Precision : %.4f'%(RF_f1_test, RF_recall_test, RF_precision_test))\n",
    "            print(confusion_matrix(inner_val_y, RF_pred_class))\n",
    "                \n",
    "    # Random Forest on logo_test data\n",
    "    clf = RandomForestClassifier(n_estimators = 500, random_state = SEED, verbose = 0, bootstrap=True)\n",
    "    clf.fit(outer_train_X, outer_train_y)\n",
    "        \n",
    "    RF_pred_class = clf.predict(outer_test_X)\n",
    "    RF_preds = clf.predict_proba(outer_test_X)\n",
    "\n",
    "    RF_AUC_test_score = roc_auc_score(outer_test_y, RF_preds[:,1])\n",
    "    RF_f1_test = f1_score(outer_test_y, RF_pred_class)\n",
    "    RF_recall_test = recall_score(outer_test_y, RF_pred_class)\n",
    "    RF_precision_test = precision_score(outer_test_y, RF_pred_class)\n",
    "    \n",
    "    #Specificity and sensitivity\n",
    "    tn, fp, fn, tp = confusion_matrix(outer_test_y, RF_pred_class).ravel()\n",
    "    specificity = tn / (tn + fp)\n",
    "    sensitivity = tp / (tp + fn)\n",
    "    \n",
    "    #Adding auc to dataframe\n",
    "    AUC_scores.at[unique_groups[i], unique_groups[i]] = RF_AUC_test_score\n",
    "    \n",
    "    #Getting AUCROC data\n",
    "    fpr, tpr, _ = roc_curve(outer_test_y,  RF_preds[:,1])\n",
    "\n",
    "    #Adding roc curve values to dataframe\n",
    "    ROC_curve.at[unique_groups[i], \"fpr\"] = \",\".join(str(x) for x in fpr)\n",
    "    ROC_curve.at[unique_groups[i], \"tpr\"] = \",\".join(str(x) for x in tpr)\n",
    "    ROC_curve.at[unique_groups[i], \"auc\"] = RF_AUC_test_score\n",
    "    \n",
    "\n",
    "    if show_fold_stats:\n",
    "        print('-' * 80)\n",
    "        print('-' * 80)\n",
    "        print('Group : %s'%(grp+1),unique_groups[i])\n",
    "        print('ROC AUC score for RandomForest model, LOGO test set: %.4f'%RF_AUC_test_score)\n",
    "        print('F1 : %.4f, Recall : %.4f , Precision : %.4f'%(RF_f1_test, RF_recall_test, RF_precision_test))\n",
    "        print(confusion_matrix(outer_test_y, RF_pred_class))\n",
    "        print('-' * 80)\n",
    "        print('-' * 80)\n",
    "    grp += 1\n",
    "        "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "id": "401268e7",
   "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>AT-CRC_species</th>\n",
       "      <th>CN-CRC_species</th>\n",
       "      <th>DE-CRC_species</th>\n",
       "      <th>FR-CRC_species</th>\n",
       "      <th>IT-CRC-2_species</th>\n",
       "      <th>IT-CRC_species</th>\n",
       "      <th>JP-CRC_species</th>\n",
       "      <th>US-CRC_species</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>AT-CRC_species</th>\n",
       "      <td>0.868703</td>\n",
       "      <td>0.894645</td>\n",
       "      <td>0.816111</td>\n",
       "      <td>0.810702</td>\n",
       "      <td>0.738839</td>\n",
       "      <td>0.782328</td>\n",
       "      <td>0.83</td>\n",
       "      <td>0.753883</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CN-CRC_species</th>\n",
       "      <td>0.874224</td>\n",
       "      <td>0.901527</td>\n",
       "      <td>0.786806</td>\n",
       "      <td>0.817352</td>\n",
       "      <td>0.759487</td>\n",
       "      <td>0.71408</td>\n",
       "      <td>0.856563</td>\n",
       "      <td>0.727441</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DE-CRC_species</th>\n",
       "      <td>0.898896</td>\n",
       "      <td>0.878003</td>\n",
       "      <td>0.814028</td>\n",
       "      <td>0.817198</td>\n",
       "      <td>0.772321</td>\n",
       "      <td>0.811782</td>\n",
       "      <td>0.877188</td>\n",
       "      <td>0.67622</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FR-CRC_species</th>\n",
       "      <td>0.858523</td>\n",
       "      <td>0.886011</td>\n",
       "      <td>0.811806</td>\n",
       "      <td>0.818126</td>\n",
       "      <td>0.788504</td>\n",
       "      <td>0.828305</td>\n",
       "      <td>0.86125</td>\n",
       "      <td>0.725407</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IT-CRC-2_species</th>\n",
       "      <td>0.864389</td>\n",
       "      <td>0.894269</td>\n",
       "      <td>0.824167</td>\n",
       "      <td>0.806681</td>\n",
       "      <td>0.758371</td>\n",
       "      <td>0.778017</td>\n",
       "      <td>0.840313</td>\n",
       "      <td>0.726886</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IT-CRC_species</th>\n",
       "      <td>0.863699</td>\n",
       "      <td>0.876752</td>\n",
       "      <td>0.828472</td>\n",
       "      <td>0.817043</td>\n",
       "      <td>0.773996</td>\n",
       "      <td>0.788793</td>\n",
       "      <td>0.850625</td>\n",
       "      <td>0.732249</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>JP-CRC_species</th>\n",
       "      <td>0.858178</td>\n",
       "      <td>0.897898</td>\n",
       "      <td>0.821389</td>\n",
       "      <td>0.826322</td>\n",
       "      <td>0.761161</td>\n",
       "      <td>0.77227</td>\n",
       "      <td>0.869375</td>\n",
       "      <td>0.73983</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>US-CRC_species</th>\n",
       "      <td>0.874224</td>\n",
       "      <td>0.873874</td>\n",
       "      <td>0.760972</td>\n",
       "      <td>0.822301</td>\n",
       "      <td>0.767299</td>\n",
       "      <td>0.817529</td>\n",
       "      <td>0.855937</td>\n",
       "      <td>0.729475</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                 AT-CRC_species CN-CRC_species DE-CRC_species FR-CRC_species  \\\n",
       "AT-CRC_species         0.868703       0.894645       0.816111       0.810702   \n",
       "CN-CRC_species         0.874224       0.901527       0.786806       0.817352   \n",
       "DE-CRC_species         0.898896       0.878003       0.814028       0.817198   \n",
       "FR-CRC_species         0.858523       0.886011       0.811806       0.818126   \n",
       "IT-CRC-2_species       0.864389       0.894269       0.824167       0.806681   \n",
       "IT-CRC_species         0.863699       0.876752       0.828472       0.817043   \n",
       "JP-CRC_species         0.858178       0.897898       0.821389       0.826322   \n",
       "US-CRC_species         0.874224       0.873874       0.760972       0.822301   \n",
       "\n",
       "                 IT-CRC-2_species IT-CRC_species JP-CRC_species US-CRC_species  \n",
       "AT-CRC_species           0.738839       0.782328           0.83       0.753883  \n",
       "CN-CRC_species           0.759487        0.71408       0.856563       0.727441  \n",
       "DE-CRC_species           0.772321       0.811782       0.877188        0.67622  \n",
       "FR-CRC_species           0.788504       0.828305        0.86125       0.725407  \n",
       "IT-CRC-2_species         0.758371       0.778017       0.840313       0.726886  \n",
       "IT-CRC_species           0.773996       0.788793       0.850625       0.732249  \n",
       "JP-CRC_species           0.761161        0.77227       0.869375        0.73983  \n",
       "US-CRC_species           0.767299       0.817529       0.855937       0.729475  "
      ]
     },
     "execution_count": 22,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "AUC_scores"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "id": "52a5b3bd",
   "metadata": {},
   "outputs": [],
   "source": [
    "import seaborn as sns"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "id": "b6feff72",
   "metadata": {},
   "outputs": [],
   "source": [
    "AUC_scores = AUC_scores.apply(pd.to_numeric, errors='coerce')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "id": "d04d8142",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 864x720 with 2 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(12,10))\n",
    "sns.set_palette(\"husl\", 8)\n",
    "ax = sns.heatmap(AUC_scores, annot = True, cmap=\"crest\", vmin=0.6, vmax=1.0, annot_kws={\"fontsize\":14})\n",
    "ax.set(xlabel=\"Validation Datasets\", ylabel=\"Test Datasets\")\n",
    "ax.tick_params(axis='x', rotation=45)\n",
    "plt.savefig(\"plots/LOSO_CLR_1.jpg\",dpi=300)"
   ]
  },
  {
   "cell_type": "markdown",
   "id": "30cc75f1",
   "metadata": {},
   "source": [
    "# ROC curves"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "id": "e9ecc1a3",
   "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>fpr</th>\n",
       "      <th>tpr</th>\n",
       "      <th>auc</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>AT-CRC_species</th>\n",
       "      <td>0.0,0.0,0.0,0.021739130434782608,0.02173913043...</td>\n",
       "      <td>0.0,0.015873015873015872,0.12698412698412698,0...</td>\n",
       "      <td>0.868703</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>CN-CRC_species</th>\n",
       "      <td>0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0135135135135135...</td>\n",
       "      <td>0.0,0.018518518518518517,0.1111111111111111,0....</td>\n",
       "      <td>0.901527</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>DE-CRC_species</th>\n",
       "      <td>0.0,0.0,0.0,0.016666666666666666,0.01666666666...</td>\n",
       "      <td>0.0,0.016666666666666666,0.08333333333333333,0...</td>\n",
       "      <td>0.814028</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>FR-CRC_species</th>\n",
       "      <td>0.0,0.0,0.0,0.018867924528301886,0.01886792452...</td>\n",
       "      <td>0.0,0.01639344262295082,0.06557377049180328,0....</td>\n",
       "      <td>0.818126</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IT-CRC-2_species</th>\n",
       "      <td>0.0,0.03125,0.03125,0.0625,0.0625,0.0625,0.062...</td>\n",
       "      <td>0.0,0.0,0.03571428571428571,0.0357142857142857...</td>\n",
       "      <td>0.758371</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>IT-CRC_species</th>\n",
       "      <td>0.0,0.0,0.0,0.034482758620689655,0.03448275862...</td>\n",
       "      <td>0.0,0.041666666666666664,0.16666666666666666,0...</td>\n",
       "      <td>0.788793</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>JP-CRC_species</th>\n",
       "      <td>0.0,0.0,0.0,0.025,0.025,0.025,0.025,0.025,0.02...</td>\n",
       "      <td>0.0,0.025,0.125,0.15,0.25,0.3,0.375,0.45,0.5,0...</td>\n",
       "      <td>0.869375</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>US-CRC_species</th>\n",
       "      <td>0.0,0.0,0.0,0.0,0.019230769230769232,0.0192307...</td>\n",
       "      <td>0.0,0.019230769230769232,0.07692307692307693,0...</td>\n",
       "      <td>0.729475</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                                fpr  \\\n",
       "AT-CRC_species    0.0,0.0,0.0,0.021739130434782608,0.02173913043...   \n",
       "CN-CRC_species    0.0,0.0,0.0,0.0,0.0,0.0,0.0,0.0135135135135135...   \n",
       "DE-CRC_species    0.0,0.0,0.0,0.016666666666666666,0.01666666666...   \n",
       "FR-CRC_species    0.0,0.0,0.0,0.018867924528301886,0.01886792452...   \n",
       "IT-CRC-2_species  0.0,0.03125,0.03125,0.0625,0.0625,0.0625,0.062...   \n",
       "IT-CRC_species    0.0,0.0,0.0,0.034482758620689655,0.03448275862...   \n",
       "JP-CRC_species    0.0,0.0,0.0,0.025,0.025,0.025,0.025,0.025,0.02...   \n",
       "US-CRC_species    0.0,0.0,0.0,0.0,0.019230769230769232,0.0192307...   \n",
       "\n",
       "                                                                tpr       auc  \n",
       "AT-CRC_species    0.0,0.015873015873015872,0.12698412698412698,0...  0.868703  \n",
       "CN-CRC_species    0.0,0.018518518518518517,0.1111111111111111,0....  0.901527  \n",
       "DE-CRC_species    0.0,0.016666666666666666,0.08333333333333333,0...  0.814028  \n",
       "FR-CRC_species    0.0,0.01639344262295082,0.06557377049180328,0....  0.818126  \n",
       "IT-CRC-2_species  0.0,0.0,0.03571428571428571,0.0357142857142857...  0.758371  \n",
       "IT-CRC_species    0.0,0.041666666666666664,0.16666666666666666,0...  0.788793  \n",
       "JP-CRC_species    0.0,0.025,0.125,0.15,0.25,0.3,0.375,0.45,0.5,0...  0.869375  \n",
       "US-CRC_species    0.0,0.019230769230769232,0.07692307692307693,0...  0.729475  "
      ]
     },
     "execution_count": 26,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "ROC_curve"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 27,
   "id": "6e4207df",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "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\n",
      "text/plain": [
       "<Figure size 720x720 with 1 Axes>"
      ]
     },
     "metadata": {
      "needs_background": "light"
     },
     "output_type": "display_data"
    }
   ],
   "source": [
    "fig = plt.figure(figsize=(10,10))\n",
    "\n",
    "for i in ROC_curve.index:\n",
    "    plt.plot([float(s) for s in ROC_curve.loc[i]['fpr'].split(\",\")], \n",
    "             [float(s) for s in ROC_curve.loc[i]['tpr'].split(\",\")], \n",
    "             label=\"{}, AUC={:.2f}\".format(i, ROC_curve.loc[i]['auc']))\n",
    "\n",
    "plt.plot([0,1], [0,1], color='orange', linestyle='--')\n",
    "\n",
    "plt.xticks(np.arange(0.0, 1.1, step=0.1))\n",
    "plt.xlabel(\"Flase Positive Rate\")#, fontsize=15)\n",
    "\n",
    "plt.yticks(np.arange(0.0, 1.1, step=0.1))\n",
    "plt.ylabel(\"True Positive Rate\")#, fontsize=15)\n",
    "\n",
    "plt.legend(prop={'size':10}, loc='lower right')\n",
    "\n",
    "plt.savefig(\"plots/LOSO_CLR_2.jpg\",dpi=300)\n",
    "plt.show()"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "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": "3.8.8"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}