titanic_tutorial.ipynb

{"metadata":{"kernelspec":{"language":"python","display_name":"Python 3","name":"python3"},"language_info":{"name":"python","version":"3.7.10","mimetype":"text/x-python","codemirror_mode":{"name":"ipython","version":3},"pygments_lexer":"ipython3","nbconvert_exporter":"python","file_extension":".py"}},"nbformat_minor":4,"nbformat":4,"cells":[{"cell_type":"code","source":"# This Python 3 environment comes with many helpful analytics libraries installed\n# It is defined by the kaggle/python Docker image: https://github.com/kaggle/docker-python\n# For example, here's several helpful packages to load\n\nimport numpy as np # linear algebra\nimport pandas as pd # data processing, CSV file I/O (e.g. pd.read_csv)\n\n# Input data files are available in the read-only \"../input/\" directory\n# For example, running this (by clicking run or pressing Shift+Enter) will list all files under the input directory\n\nimport os\nfor dirname, _, filenames in os.walk('/kaggle/input'):\n    for filename in filenames:\n        print(os.path.join(dirname, filename))\n\n# You can write up to 20GB to the current directory (/kaggle/working/) that gets preserved as output when you create a version using \"Save & Run All\" \n# You can also write temporary files to /kaggle/temp/, but they won't be saved outside of the current session","metadata":{"_uuid":"8f2839f25d086af736a60e9eeb907d3b93b6e0e5","_cell_guid":"b1076dfc-b9ad-4769-8c92-a6c4dae69d19","execution":{"iopub.status.busy":"2021-10-13T05:50:38.751038Z","iopub.execute_input":"2021-10-13T05:50:38.751847Z","iopub.status.idle":"2021-10-13T05:50:38.777508Z","shell.execute_reply.started":"2021-10-13T05:50:38.751758Z","shell.execute_reply":"2021-10-13T05:50:38.776939Z"},"trusted":true},"execution_count":1,"outputs":[]},{"cell_type":"code","source":"import matplotlib.pyplot as plt\nimport seaborn as sns\n\nplt.style.use('seaborn')\nsns.set(font_scale=2.5) #matplot의 기본 scheme 대신 seaborn scheme 사용\nimport missingno as msno\n\n#ignore warnings\nimport warnings\nwarnings.filterwarnings('ignore')\n\n%matplotlib inline","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:38.778591Z","iopub.execute_input":"2021-10-13T05:50:38.779350Z","iopub.status.idle":"2021-10-13T05:50:39.633623Z","shell.execute_reply.started":"2021-10-13T05:50:38.779306Z","shell.execute_reply":"2021-10-13T05:50:39.633025Z"},"trusted":true},"execution_count":2,"outputs":[]},{"cell_type":"markdown","source":"1. Dataset 확인","metadata":{}},{"cell_type":"code","source":"df_train = pd.read_csv('../input/titanic/train.csv')\ndf_test = pd.read_csv('../input/titanic/test.csv')\n\ndf_train.head()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:39.634619Z","iopub.execute_input":"2021-10-13T05:50:39.634925Z","iopub.status.idle":"2021-10-13T05:50:39.681430Z","shell.execute_reply.started":"2021-10-13T05:50:39.634900Z","shell.execute_reply":"2021-10-13T05:50:39.680843Z"},"trusted":true},"execution_count":3,"outputs":[]},{"cell_type":"code","source":"#feature들의 통계치 확인\ndf_train.describe()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:39.682904Z","iopub.execute_input":"2021-10-13T05:50:39.683242Z","iopub.status.idle":"2021-10-13T05:50:39.720213Z","shell.execute_reply.started":"2021-10-13T05:50:39.683217Z","shell.execute_reply":"2021-10-13T05:50:39.719388Z"},"trusted":true},"execution_count":4,"outputs":[]},{"cell_type":"code","source":"df_test.describe()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:39.721395Z","iopub.execute_input":"2021-10-13T05:50:39.721594Z","iopub.status.idle":"2021-10-13T05:50:39.749901Z","shell.execute_reply.started":"2021-10-13T05:50:39.721571Z","shell.execute_reply":"2021-10-13T05:50:39.749010Z"},"trusted":true},"execution_count":5,"outputs":[]},{"cell_type":"markdown","source":"PassengerId의 count와 다른 수의 feature들이 보임 -> null data 확인!\n\n1.1 Check Null Data","metadata":{}},{"cell_type":"code","source":"# 1. null data 비율\n\nfor col in df_train.columns:\n    msg = 'column: {:>10}\\t Percent of NaN value: {:.2f}%'.format(col, 100 * (df_train[col].isnull().sum() / df_train[col].shape[0]))\n    \n    print(msg)","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:39.751333Z","iopub.execute_input":"2021-10-13T05:50:39.751874Z","iopub.status.idle":"2021-10-13T05:50:39.764656Z","shell.execute_reply.started":"2021-10-13T05:50:39.751832Z","shell.execute_reply":"2021-10-13T05:50:39.763704Z"},"trusted":true},"execution_count":6,"outputs":[]},{"cell_type":"code","source":"for col in df_test.columns:\n    msg = 'column: {:>10}\\t Percent of NaN value: {:.2f}%'.format(col, 100 * (df_test[col].isnull().sum() / df_test[col].shape[0]))\n    print(msg)","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:39.765675Z","iopub.execute_input":"2021-10-13T05:50:39.765893Z","iopub.status.idle":"2021-10-13T05:50:39.776289Z","shell.execute_reply.started":"2021-10-13T05:50:39.765869Z","shell.execute_reply":"2021-10-13T05:50:39.775686Z"},"trusted":true},"execution_count":7,"outputs":[]},{"cell_type":"code","source":"# 2. MSNO library 이용한 시각화\n\nmsno.matrix(df=df_train.iloc[:,:], figsize=(8,8), color=(0.8,0.5,0.2))","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:39.777100Z","iopub.execute_input":"2021-10-13T05:50:39.777321Z","iopub.status.idle":"2021-10-13T05:50:40.142384Z","shell.execute_reply.started":"2021-10-13T05:50:39.777296Z","shell.execute_reply":"2021-10-13T05:50:40.141437Z"},"trusted":true},"execution_count":8,"outputs":[]},{"cell_type":"code","source":"msno.bar(df=df_train.iloc[:, :], figsize=(8, 8), color=(0.8, 0.5, 0.2))","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:40.143691Z","iopub.execute_input":"2021-10-13T05:50:40.144020Z","iopub.status.idle":"2021-10-13T05:50:40.966813Z","shell.execute_reply.started":"2021-10-13T05:50:40.143977Z","shell.execute_reply":"2021-10-13T05:50:40.965796Z"},"trusted":true},"execution_count":9,"outputs":[]},{"cell_type":"markdown","source":"1.2 Target Label Distribution 확인\n\nTraining Dataset이 균일하게 분포되지 않고 치우쳐 있다면, 제대로 분류하지 못하게 된다.\n\nBinary Classification의 경우 0이 99%, 1이 1% 로 이루어진 데이터로 학습했다면, 다 0이라고 판단해도 정확도가 99%가 나올 것이다","metadata":{}},{"cell_type":"code","source":"f, ax = plt.subplots(1, 2, figsize=(18, 8))\n\ndf_train['Survived'].value_counts().plot.pie(explode=[0, 0.1], autopct='%1.1f%%', ax=ax[0], shadow=True)\nax[0].set_title('Pie plot - Survived')\nax[0].set_ylabel('')\nsns.countplot('Survived', data=df_train, ax=ax[1])\nax[1].set_title('Count plot - Survived')\n\nplt.show()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"2. Exploratory Data Analysis\n\nfeature에 따른 생존률의 차이 확인!","metadata":{}},{"cell_type":"markdown","source":"2.1 Pclass","metadata":{}},{"cell_type":"code","source":"df_train[['Pclass', 'Survived']].groupby(['Pclass'], as_index=True).count()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:40.970587Z","iopub.execute_input":"2021-10-13T05:50:40.970997Z","iopub.status.idle":"2021-10-13T05:50:40.985398Z","shell.execute_reply.started":"2021-10-13T05:50:40.970944Z","shell.execute_reply":"2021-10-13T05:50:40.984485Z"},"trusted":true},"execution_count":10,"outputs":[]},{"cell_type":"code","source":"df_train[['Pclass', 'Survived']].groupby(['Pclass'], as_index=True).sum()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:40.986726Z","iopub.execute_input":"2021-10-13T05:50:40.987574Z","iopub.status.idle":"2021-10-13T05:50:41.002568Z","shell.execute_reply.started":"2021-10-13T05:50:40.987531Z","shell.execute_reply":"2021-10-13T05:50:41.001843Z"},"trusted":true},"execution_count":11,"outputs":[]},{"cell_type":"code","source":"# pandas의 crosstab 표\npd.crosstab(df_train['Pclass'], df_train['Survived'], margins=True).style.background_gradient(cmap='summer_r')","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:41.003753Z","iopub.execute_input":"2021-10-13T05:50:41.004548Z","iopub.status.idle":"2021-10-13T05:50:41.128280Z","shell.execute_reply.started":"2021-10-13T05:50:41.004507Z","shell.execute_reply":"2021-10-13T05:50:41.127460Z"},"trusted":true},"execution_count":12,"outputs":[]},{"cell_type":"code","source":"# 각 클래스 (Pclass) 별 생존률\ndf_train[['Pclass', 'Survived']].groupby(['Pclass'], as_index=True).mean().sort_values(by='Survived', ascending=False)\\\n.plot.bar()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:41.129611Z","iopub.execute_input":"2021-10-13T05:50:41.130313Z","iopub.status.idle":"2021-10-13T05:50:41.308372Z","shell.execute_reply.started":"2021-10-13T05:50:41.130282Z","shell.execute_reply":"2021-10-13T05:50:41.307415Z"},"trusted":true},"execution_count":13,"outputs":[]},{"cell_type":"code","source":"y_position = 1.02\nf, ax = plt.subplots(1, 2, figsize=(18, 8))\ndf_train['Pclass'].value_counts().plot.bar(color=['#CD7F32','#FFDF00','#D3D3D3'], ax=ax[0])\nax[0].set_title('Number of Passengers By Pclass', y=y_position)\nax[0].set_ylabel('Count')\nsns.countplot('Pclass', hue='Survived', data=df_train, ax=ax[1])\nax[1].set_title('Pclass: Survived vs Dead', y=y_position)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:41.309762Z","iopub.execute_input":"2021-10-13T05:50:41.310140Z","iopub.status.idle":"2021-10-13T05:50:41.755320Z","shell.execute_reply.started":"2021-10-13T05:50:41.310097Z","shell.execute_reply":"2021-10-13T05:50:41.754392Z"},"trusted":true},"execution_count":14,"outputs":[]},{"cell_type":"markdown","source":"2.2 Sex","metadata":{}},{"cell_type":"code","source":"# seaborn의 countplot 이용\n\nf, ax = plt.subplots(1, 2, figsize=(18, 8))\ndf_train[['Sex', 'Survived']].groupby(['Sex'], as_index=True).mean().plot.bar(ax=ax[0])\nax[0].set_title('Survived vs Sex')\nsns.countplot('Sex', hue='Survived', data=df_train, ax=ax[1])\nax[1].set_title('Sex: Survived vs Dead')\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:41.756612Z","iopub.execute_input":"2021-10-13T05:50:41.756888Z","iopub.status.idle":"2021-10-13T05:50:42.140941Z","shell.execute_reply.started":"2021-10-13T05:50:41.756852Z","shell.execute_reply":"2021-10-13T05:50:42.140133Z"},"trusted":true},"execution_count":15,"outputs":[]},{"cell_type":"code","source":"df_train[['Sex', 'Survived']].groupby(['Sex'], as_index=False).mean().sort_values(by='Survived', ascending=False)","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:42.142074Z","iopub.execute_input":"2021-10-13T05:50:42.142330Z","iopub.status.idle":"2021-10-13T05:50:42.157917Z","shell.execute_reply.started":"2021-10-13T05:50:42.142303Z","shell.execute_reply":"2021-10-13T05:50:42.157104Z"},"trusted":true},"execution_count":16,"outputs":[]},{"cell_type":"code","source":"pd.crosstab(df_train['Sex'], df_train['Survived'], margins=True).style.background_gradient(cmap='summer_r')","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:42.158889Z","iopub.execute_input":"2021-10-13T05:50:42.159097Z","iopub.status.idle":"2021-10-13T05:50:42.209359Z","shell.execute_reply.started":"2021-10-13T05:50:42.159073Z","shell.execute_reply":"2021-10-13T05:50:42.208659Z"},"trusted":true},"execution_count":17,"outputs":[]},{"cell_type":"markdown","source":"2.3 Pclass + Sex","metadata":{}},{"cell_type":"code","source":"# seaborn의 factorplot 이용\n# hue='Sex' -> 색상으로 구분\n\nsns.factorplot('Pclass', 'Survived', hue='Sex', data=df_train, size=6, aspect=1.5)","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:42.210405Z","iopub.execute_input":"2021-10-13T05:50:42.210615Z","iopub.status.idle":"2021-10-13T05:50:42.877044Z","shell.execute_reply.started":"2021-10-13T05:50:42.210584Z","shell.execute_reply":"2021-10-13T05:50:42.876245Z"},"trusted":true},"execution_count":18,"outputs":[]},{"cell_type":"code","source":"# column\n\nsns.factorplot(x='Sex', y='Survived', col='Pclass', data=df_train, satureation=.5, size=9, aspect=1)","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:42.878421Z","iopub.execute_input":"2021-10-13T05:50:42.878697Z","iopub.status.idle":"2021-10-13T05:50:43.873802Z","shell.execute_reply.started":"2021-10-13T05:50:42.878668Z","shell.execute_reply":"2021-10-13T05:50:43.873266Z"},"trusted":true},"execution_count":19,"outputs":[]},{"cell_type":"markdown","source":"2.4 Age","metadata":{}},{"cell_type":"code","source":"print(\"The oldest : {:.1f}\".format(df_train['Age'].max()))\nprint(\"The youngest : {:.1f}\".format(df_train['Age'].min()))\nprint(\"Average age : {:.1f}\".format(df_train['Age'].mean()))","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:43.874661Z","iopub.execute_input":"2021-10-13T05:50:43.875305Z","iopub.status.idle":"2021-10-13T05:50:43.880994Z","shell.execute_reply.started":"2021-10-13T05:50:43.875277Z","shell.execute_reply":"2021-10-13T05:50:43.880243Z"},"trusted":true},"execution_count":20,"outputs":[]},{"cell_type":"code","source":"# Survive age distribution\n\nfig, ax = plt.subplots(1, 1, figsize=(9, 5))\nsns.kdeplot(df_train[df_train['Survived'] == 1]['Age'], ax=ax)\nsns.kdeplot(df_train[df_train['Survived'] == 0]['Age'], ax=ax)\nplt.legend(['Survived == 1', 'Survived == 0'])\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:43.882106Z","iopub.execute_input":"2021-10-13T05:50:43.882535Z","iopub.status.idle":"2021-10-13T05:50:44.186939Z","shell.execute_reply.started":"2021-10-13T05:50:43.882506Z","shell.execute_reply":"2021-10-13T05:50:44.186128Z"},"trusted":true},"execution_count":21,"outputs":[]},{"cell_type":"code","source":"# Age distribution within classes\n\nplt.figure(figsize=(8,6))\ndf_train['Age'][df_train['Pclass']==1].plot(kind='kde')\ndf_train['Age'][df_train['Pclass'] == 2].plot(kind='kde')\ndf_train['Age'][df_train['Pclass'] == 3].plot(kind='kde')\n\nplt.xlabel('Age')\nplt.title('Age Distribution within classes')\nplt.legend(['1st Class', '2nd Class', '3rd Class'])","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:44.188352Z","iopub.execute_input":"2021-10-13T05:50:44.188648Z","iopub.status.idle":"2021-10-13T05:50:44.555583Z","shell.execute_reply.started":"2021-10-13T05:50:44.188607Z","shell.execute_reply":"2021-10-13T05:50:44.554758Z"},"trusted":true},"execution_count":22,"outputs":[]},{"cell_type":"code","source":"cummulative_survival_ratio = []\nfor i in range(1, 80):\n    cummulative_survival_ratio.append(df_train[df_train['Age']<i]['Survived'].sum() / len(df_train[df_train['Age']<i]['Survived']))\n    \nplt.figure(figsize=(7,7))\nplt.plot(cummulative_survival_ratio)\nplt.title('Survival rate change depending on range of age', y=1.02)\nplt.ylabel('Survival rate')\nplt.xlabel('Range of age(0~x)')\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:44.556839Z","iopub.execute_input":"2021-10-13T05:50:44.557316Z","iopub.status.idle":"2021-10-13T05:50:44.846695Z","shell.execute_reply.started":"2021-10-13T05:50:44.557273Z","shell.execute_reply":"2021-10-13T05:50:44.845934Z"},"trusted":true},"execution_count":23,"outputs":[]},{"cell_type":"markdown","source":"2.5 Pclass + Sex + Age","metadata":{}},{"cell_type":"code","source":"f,ax=plt.subplots(1,2,figsize=(18,8))\nsns.violinplot(\"Pclass\",\"Age\", hue=\"Survived\", data=df_train, scale='count', split=True,ax=ax[0])\nax[0].set_title('Pclass and Age vs Survived')\nax[0].set_yticks(range(0,110,10))\nsns.violinplot(\"Sex\",\"Age\", hue=\"Survived\", data=df_train, scale='count', split=True,ax=ax[1])\nax[1].set_title('Sex and Age vs Survived')\nax[1].set_yticks(range(0,110,10))\nplt.show()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"2.6 Embarked","metadata":{}},{"cell_type":"code","source":"f, ax = plt.subplots(1, 1, figsize=(7, 7))\ndf_train[['Embarked', 'Survived']].groupby(['Embarked'], as_index=True).mean().sort_values(by='Survived', ascending=False).plot.bar(ax=ax)","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:44.847931Z","iopub.execute_input":"2021-10-13T05:50:44.848241Z","iopub.status.idle":"2021-10-13T05:50:45.094146Z","shell.execute_reply.started":"2021-10-13T05:50:44.848201Z","shell.execute_reply":"2021-10-13T05:50:45.093310Z"},"trusted":true},"execution_count":24,"outputs":[]},{"cell_type":"code","source":"f,ax=plt.subplots(2, 2, figsize=(20,15))\n\nsns.countplot('Embarked', data=df_train, ax=ax[0,0])\nax[0,0].set_title('(1) No. Of Passengers Boarded')\n\nsns.countplot('Embarked', hue='Sex', data=df_train, ax=ax[0,1])\nax[0,1].set_title('(2) Male-Female Split for Embarked')\n\nsns.countplot('Embarked', hue='Survived', data=df_train, ax=ax[1,0])\nax[1,0].set_title('(3) Embarked vs Survived')\n\nsns.countplot('Embarked', hue='Pclass', data=df_train, ax=ax[1,1])\nax[1,1].set_title('(4) Embarked vs Pclass')\n\nplt.subplots_adjust(wspace=0.2, hspace=0.5)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:45.095368Z","iopub.execute_input":"2021-10-13T05:50:45.095605Z","iopub.status.idle":"2021-10-13T05:50:45.815010Z","shell.execute_reply.started":"2021-10-13T05:50:45.095577Z","shell.execute_reply":"2021-10-13T05:50:45.814207Z"},"trusted":true},"execution_count":25,"outputs":[]},{"cell_type":"markdown","source":"2.7 Family - SibSp + Parch","metadata":{}},{"cell_type":"code","source":"df_train['FamilySize'] = df_train['SibSp'] + df_train['Parch'] + 1\ndf_test['FamilySize'] = df_test['SibSp'] + df_test['Parch'] + 1","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:45.816052Z","iopub.execute_input":"2021-10-13T05:50:45.816293Z","iopub.status.idle":"2021-10-13T05:50:45.823712Z","shell.execute_reply.started":"2021-10-13T05:50:45.816266Z","shell.execute_reply":"2021-10-13T05:50:45.822772Z"},"trusted":true},"execution_count":26,"outputs":[]},{"cell_type":"code","source":"f, ax = plt.subplots(1, 3, figsize=(40,10))\n\nsns.countplot('FamilySize', data=df_train, ax=ax[0])\nax[0].set_title('(1) No. Of Passengers Boarded', y=1.02)\n\nsns.countplot('FamilySize', hue='Survived', data=df_train, ax=ax[1])\nax[1].set_title('(2) Survived countplot depending on FamilySize',  y=1.02)\n\ndf_train[['FamilySize', 'Survived']].groupby(['FamilySize'], as_index=True).mean().sort_values(by='Survived', ascending=False).plot.bar(ax=ax[2])\nax[2].set_title('(3) Survived rate depending on FamilySize',  y=1.02)\n\nplt.subplots_adjust(wspace=0.2, hspace=0.5)\nplt.show()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:50:45.824919Z","iopub.execute_input":"2021-10-13T05:50:45.825127Z","iopub.status.idle":"2021-10-13T05:50:46.532102Z","shell.execute_reply.started":"2021-10-13T05:50:45.825103Z","shell.execute_reply":"2021-10-13T05:50:46.531375Z"},"trusted":true},"execution_count":27,"outputs":[]},{"cell_type":"markdown","source":"2.8 Fare","metadata":{}},{"cell_type":"code","source":"fig, ax = plt.subplots(1, 1, figsize=(8, 8))\ng = sns.distplot(df_train['Fare'], color='b', label='Skewness : {:.2f}'.format(df_train['Fare'].skew()), ax=ax)\ng = g.legend(loc='best')","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:52:22.829975Z","iopub.execute_input":"2021-10-13T05:52:22.830291Z","iopub.status.idle":"2021-10-13T05:52:23.260661Z","shell.execute_reply.started":"2021-10-13T05:52:22.830262Z","shell.execute_reply":"2021-10-13T05:52:23.259742Z"},"trusted":true},"execution_count":28,"outputs":[]},{"cell_type":"markdown","source":"-> distribution이 비대칭이다\n\n-> model이 이 데이터를 학습한다면, outlier에 대해서 잘못 학습할 수 있다\n\n-> outlier의 영향을 줄이기 위해 log를 취해주자!\n\n\n* feature engineering\n\n    모델을 학습시키기 위해, 그리고 그 모델의 성능을 높이기 위해 feature 들에 여러 조작을 가하거나 새로운 feature를 추가하는 것","metadata":{}},{"cell_type":"code","source":"# Fare columns 데이터에 log값 취해주기\n\ndf_test.loc[df_test.Fare.isnull(), 'Fare'] = df_test['Fare'].mean() # testset 에 있는 nan value 를 평균값으로 치환\n\ndf_train['Fare'] = df_train['Fare'].map(lambda i: np.log(i) if i > 0 else 0)\ndf_test['Fare'] = df_test['Fare'].map(lambda i: np.log(i) if i > 0 else 0)","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:55:56.678834Z","iopub.execute_input":"2021-10-13T05:55:56.679141Z","iopub.status.idle":"2021-10-13T05:55:56.691106Z","shell.execute_reply.started":"2021-10-13T05:55:56.679111Z","shell.execute_reply":"2021-10-13T05:55:56.690258Z"},"trusted":true},"execution_count":29,"outputs":[]},{"cell_type":"code","source":"fig, ax = plt.subplots(1, 1, figsize=(8, 8))\ng = sns.distplot(df_train['Fare'], color='b', label='Skewness : {:.2f}'.format(df_train['Fare'].skew()), ax=ax)\ng = g.legend(loc='best')","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:56:02.477155Z","iopub.execute_input":"2021-10-13T05:56:02.477771Z","iopub.status.idle":"2021-10-13T05:56:03.032022Z","shell.execute_reply.started":"2021-10-13T05:56:02.477732Z","shell.execute_reply":"2021-10-13T05:56:03.031169Z"},"trusted":true},"execution_count":30,"outputs":[]},{"cell_type":"markdown","source":"2.9 Cabin","metadata":{}},{"cell_type":"code","source":"df_train.head()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T05:58:00.428123Z","iopub.execute_input":"2021-10-13T05:58:00.428454Z","iopub.status.idle":"2021-10-13T05:58:00.445967Z","shell.execute_reply.started":"2021-10-13T05:58:00.428423Z","shell.execute_reply":"2021-10-13T05:58:00.445148Z"},"trusted":true},"execution_count":31,"outputs":[]},{"cell_type":"code","source":"print(\"Cabin column의 NaN 데이터 비율 : {:.1f}%\".format(100 * (df_train['Cabin'].isnull().sum() / df_train['Cabin'].shape[0])))","metadata":{"execution":{"iopub.status.busy":"2021-10-13T06:04:31.101032Z","iopub.execute_input":"2021-10-13T06:04:31.101389Z","iopub.status.idle":"2021-10-13T06:04:31.108734Z","shell.execute_reply.started":"2021-10-13T06:04:31.101355Z","shell.execute_reply":"2021-10-13T06:04:31.107919Z"},"trusted":true},"execution_count":39,"outputs":[]},{"cell_type":"markdown","source":"NaN의 비율이 77%로, 생존에 영향을 끼칠 중요한 정보를 알아내기 힘들다.","metadata":{}},{"cell_type":"markdown","source":"2.10 Ticket","metadata":{}},{"cell_type":"code","source":"df_train['Ticket'].value_counts()","metadata":{"execution":{"iopub.status.busy":"2021-10-13T06:06:07.026251Z","iopub.execute_input":"2021-10-13T06:06:07.026571Z","iopub.status.idle":"2021-10-13T06:06:07.037087Z","shell.execute_reply.started":"2021-10-13T06:06:07.026540Z","shell.execute_reply":"2021-10-13T06:06:07.036228Z"},"trusted":true},"execution_count":40,"outputs":[]},{"cell_type":"markdown","source":"3. Feature Engineering","metadata":{}},{"cell_type":"code","source":"3.1 Fill Null data","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"3.1.1 Fill Null in age using title","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train['Initial']= df_train.Name.str.extract('([A-Za-z]+)\\.') #lets extract the Salutations\n    \ndf_test['Initial']= df_test.Name.str.extract('([A-Za-z]+)\\.') #lets extract the Salutations\n\npd.crosstab(df_train['Initial'], df_train['Sex']).T.style.background_gradient(cmap='summer_r') #Checking the Initials with the Sex\n","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train['Initial'].replace(['Mlle','Mme','Ms','Dr','Major','Lady','Countess','Jonkheer','Col','Rev','Capt','Sir','Don', 'Dona'],\n                        ['Miss','Miss','Miss','Mr','Mr','Mrs','Mrs','Other','Other','Other','Mr','Mr','Mr', 'Mr'],inplace=True)\n\ndf_test['Initial'].replace(['Mlle','Mme','Ms','Dr','Major','Lady','Countess','Jonkheer','Col','Rev','Capt','Sir','Don', 'Dona'],\n                        ['Miss','Miss','Miss','Mr','Mr','Mrs','Mrs','Other','Other','Other','Mr','Mr','Mr', 'Mr'],inplace=True)\ndf_train.groupby('Initial').mean()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.groupby('Initial')['Survived'].mean().plot.bar()\n\ndf_train.groupby('Initial').mean()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.loc[(df_train.Age.isnull())&(df_train.Initial=='Mr'),'Age'] = 33\ndf_train.loc[(df_train.Age.isnull())&(df_train.Initial=='Mrs'),'Age'] = 36\ndf_train.loc[(df_train.Age.isnull())&(df_train.Initial=='Master'),'Age'] = 5\ndf_train.loc[(df_train.Age.isnull())&(df_train.Initial=='Miss'),'Age'] = 22\ndf_train.loc[(df_train.Age.isnull())&(df_train.Initial=='Other'),'Age'] = 46\n\ndf_test.loc[(df_test.Age.isnull())&(df_test.Initial=='Mr'),'Age'] = 33\ndf_test.loc[(df_test.Age.isnull())&(df_test.Initial=='Mrs'),'Age'] = 36\ndf_test.loc[(df_test.Age.isnull())&(df_test.Initial=='Master'),'Age'] = 5\ndf_test.loc[(df_test.Age.isnull())&(df_test.Initial=='Miss'),'Age'] = 22\ndf_test.loc[(df_test.Age.isnull())&(df_test.Initial=='Other'),'Age'] = 46","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"3.1.2 Fill Null in Embarked","metadata":{}},{"cell_type":"code","source":"print('Embarked has ', sum(df_train['Embarked'].isnull()), ' Null values')\ndf_train['Embarked'].fillna('S', inplace=True)","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"3.2 Change Age (continuous -> categorical)","metadata":{}},{"cell_type":"code","source":"df_train['Age_cat'] = 0\ndf_train.loc[df_train['Age'] < 10, 'Age_cat'] = 0\ndf_train.loc[(10 <= df_train['Age']) & (df_train['Age'] < 20), 'Age_cat'] = 1\ndf_train.loc[(20 <= df_train['Age']) & (df_train['Age'] < 30), 'Age_cat'] = 2\ndf_train.loc[(30 <= df_train['Age']) & (df_train['Age'] < 40), 'Age_cat'] = 3\ndf_train.loc[(40 <= df_train['Age']) & (df_train['Age'] < 50), 'Age_cat'] = 4\ndf_train.loc[(50 <= df_train['Age']) & (df_train['Age'] < 60), 'Age_cat'] = 5\ndf_train.loc[(60 <= df_train['Age']) & (df_train['Age'] < 70), 'Age_cat'] = 6\ndf_train.loc[70 <= df_train['Age'], 'Age_cat'] = 7\n\ndf_test['Age_cat'] = 0\ndf_test.loc[df_test['Age'] < 10, 'Age_cat'] = 0\ndf_test.loc[(10 <= df_test['Age']) & (df_test['Age'] < 20), 'Age_cat'] = 1\ndf_test.loc[(20 <= df_test['Age']) & (df_test['Age'] < 30), 'Age_cat'] = 2\ndf_test.loc[(30 <= df_test['Age']) & (df_test['Age'] < 40), 'Age_cat'] = 3\ndf_test.loc[(40 <= df_test['Age']) & (df_test['Age'] < 50), 'Age_cat'] = 4\ndf_test.loc[(50 <= df_test['Age']) & (df_test['Age'] < 60), 'Age_cat'] = 5\ndf_test.loc[(60 <= df_test['Age']) & (df_test['Age'] < 70), 'Age_cat'] = 6\ndf_test.loc[70 <= df_test['Age'], 'Age_cat'] = 7","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"def category_age(x):\n    if x < 10:\n        return 0\n    elif x < 20:\n        return 1\n    elif x < 30:\n        return 2\n    elif x < 40:\n        return 3\n    elif x < 50:\n        return 4\n    elif x < 60:\n        return 5\n    elif x < 70:\n        return 6\n    else:\n        return 7    \n    \ndf_train['Age_cat_2'] = df_train['Age'].apply(category_age)","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print('1번 방법, 2번 방법 둘다 같은 결과를 내면 True 줘야함 -> ', (df_train['Age_cat'] == df_train['Age_cat_2']).all())","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train.drop(['Age', 'Age_cat_2'], axis=1, inplace=True)\ndf_test.drop(['Age'], axis=1, inplace=True)","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"3.3 Change Initial, Embarked and Sex (string -> numerical)","metadata":{}},{"cell_type":"code","source":"df_train['Initial'] = df_train['Initial'].map({'Master': 0, 'Miss': 1, 'Mr': 2, 'Mrs': 3, 'Other': 4})\ndf_test['Initial'] = df_test['Initial'].map({'Master': 0, 'Miss': 1, 'Mr': 2, 'Mrs': 3, 'Other': 4})","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train['Embarked'].unique()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train['Embarked'].value_counts()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train['Embarked'] = df_train['Embarked'].map({'C': 0, 'Q': 1, 'S': 2})\ndf_test['Embarked'] = df_test['Embarked'].map({'C': 0, 'Q': 1, 'S': 2})","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train['Embarked'].isnull().any()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train['Sex'] = df_train['Sex'].map({'female': 0, 'male': 1})\ndf_test['Sex'] = df_test['Sex'].map({'female': 0, 'male': 1})","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"heatmap_data = df_train[['Survived', 'Pclass', 'Sex', 'Fare', 'Embarked', 'FamilySize', 'Initial', 'Age_cat']] \n\ncolormap = plt.cm.RdBu\nplt.figure(figsize=(14, 12))\nplt.title('Pearson Correlation of Features', y=1.05, size=15)\nsns.heatmap(heatmap_data.astype(float).corr(), linewidths=0.1, vmax=1.0,\n           square=True, cmap=colormap, linecolor='white', annot=True, annot_kws={\"size\": 16})\n\ndel heatmap_data","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"3.4 One-hot Encoding on Initial and Embarked","metadata":{}},{"cell_type":"code","source":"df_train = pd.get_dummies(df_train, columns=['Initial'], prefix='Initial')\ndf_test = pd.get_dummies(df_test, columns=['Initial'], prefix='Initial')\ndf_train.head()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"df_train = pd.get_dummies(df_train, columns=['Embarked'], prefix='Embarked')\ndf_test = pd.get_dummies(df_test, columns=['Embarked'], prefix='Embarked')","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"3.5 Drop columns","metadata":{}},{"cell_type":"code","source":"df_train.drop(['PassengerId', 'Name', 'SibSp', 'Parch', 'Ticket', 'Cabin'], axis=1, inplace=True)\ndf_test.drop(['PassengerId', 'Name',  'SibSp', 'Parch', 'Ticket', 'Cabin'], axis=1, inplace=True)\ndf_train.head()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"4. Build machine learning model and predict using the trained model","metadata":{}},{"cell_type":"code","source":"#importing all the required ML packages\nfrom sklearn.ensemble import RandomForestClassifier # 유명한 randomforestclassfier 입니다. \nfrom sklearn import metrics # 모델의 평가를 위해서 씁니다\nfrom sklearn.model_selection import train_test_split # traning set을 쉽게 나눠주는 함수입니다.","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"4.1 Preparation - Split Dataset into train, valid, and test set","metadata":{}},{"cell_type":"code","source":"X_train = df_train.drop('Survived', axis=1).values\ntarget_label = df_train['Survived'].values\nX_test = df_test.values","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"X_tr, X_vld, y_tr, y_vld = train_test_split(X_train, target_label, test_size=0.3, random_state=2018)","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"4.2 Model gerneration and prediction","metadata":{}},{"cell_type":"code","source":"model = RandomForestClassifier()\nmodel.fit(X_tr, y_tr)\nprediction = model.predict(X_vld)","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"print('총 {}명 중 {:.2f}% 정확도로 생존을 맞춤'.format(y_vld.shape[0], 100 * metrics.accuracy_score(prediction, y_vld)))","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"4.3 Feature importance","metadata":{}},{"cell_type":"code","source":"from pandas import Series\n\nfeature_importance = model.feature_importances_\nSeries_feat_imp = Series(feature_importance, index=df_test.columns)\nplt.figure(figsize=(8, 8))\nSeries_feat_imp.sort_values(ascending=True).plot.barh()\nplt.xlabel('Feature importance')\nplt.ylabel('Feature')\nplt.show()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"markdown","source":"4.4 Prediction on Test set","metadata":{}},{"cell_type":"code","source":"submission = pd.read_csv('../input/gender_submission.csv')\nsubmission.head()","metadata":{},"execution_count":null,"outputs":[]},{"cell_type":"code","source":"prediction = model.predict(X_test)\nsubmission['Survived'] = prediction\nsubmission.to_csv('./my_first_submission.csv', index=False)","metadata":{},"execution_count":null,"outputs":[]}]}