Skip to content

Commit 43d362f

Browse files
sandipanpaul21sandipanpaul21
committed
Created using Colaboratory
1 parent d07dbaa commit 43d362f

File tree

1 file changed

+148
-7
lines changed

1 file changed

+148
-7
lines changed

‎08_Logistic_Regression.ipynb

Lines changed: 148 additions & 7 deletions
Original file line numberDiff line numberDiff line change
@@ -5,7 +5,7 @@
55
"colab": {
66
"name": "08 Logistic Regression.ipynb",
77
"provenance": [],
8-
"authorship_tag": "ABX9TyOu3EEwuHsR/3JTtbVCA2NW",
8+
"authorship_tag": "ABX9TyNd8trmiuHbcgUjcCZGui3z",
99
"include_colab_link": true
1010
},
1111
"kernelspec": {
@@ -668,7 +668,7 @@
668668
"source": [
669669
"# Model Prediction\n",
670670
"\n",
671-
"print(\"Sample Prediction\")\n",
671+
"print(\"Sample Prediction of Model 1\")\n",
672672
"pred = result.predict(X_test)\n",
673673
"model_prediction = pd.DataFrame(pred.round(2),columns = ['Prediction'])\n",
674674
"model_prediction['temp'] = 'temp'\n",
@@ -724,23 +724,164 @@
724724
"metadata": {
725725
"id": "dhcskdb5BCiU",
726726
"colab_type": "code",
727-
"colab": {}
727+
"colab": {
728+
"base_uri": "https://localhost:8080/",
729+
"height": 593
730+
},
731+
"outputId": "ecdbf054-6db5-476c-a9af-d566378d8785"
728732
},
729733
"source": [
730-
""
734+
"# Lets build model with all variables\n",
735+
"X_train, X_test, y_train, y_test = train_test_split(Independent_Variable_Base_Set,Dependent_Variable,test_size = 0.3,random_state = 21)\n",
736+
"logit_model = sm.Logit(y_train,X_train)\n",
737+
"result = logit_model.fit(method='bfgs')\n",
738+
"print(result.summary2())\n",
739+
"\n",
740+
"# Model Summary\n",
741+
"\n",
742+
"print(\"Model 2 Summary\")\n",
743+
"print(\"Iteration suggests how many loop model did to perform the fit\")\n",
744+
"print(\"Iterations : 22\")\n",
745+
"r_square_2 = result.prsquared.round(2)\n",
746+
"print(\"Pseudo R Square suggests overall effect size (ideal value is close to 1)\")\n",
747+
"print(\"Model 2, MacFadden Pseudo R Square : \",r_square_2)\n",
748+
"base_model_aic_2 = result.aic.round(2)\n",
749+
"print(\"AIC compares Goodness of Fit, Lower AIC better is the Model\")\n",
750+
"print(\"Model 2, AIC :\",base_model_aic_2)\n",
751+
"base_model_bic_2 = result.bic.round(2)\n",
752+
"print(\"BIC also work same as AIC, Lower BIC better is the Model\")\n",
753+
"print(\"Model 2, BIC :\",base_model_bic_2)"
731754
],
732-
"execution_count": 10,
733-
"outputs": []
755+
"execution_count": 14,
756+
"outputs": [
757+
{
758+
"output_type": "stream",
759+
"text": [
760+
"Optimization terminated successfully.\n",
761+
" Current function value: 0.000001\n",
762+
" Iterations: 22\n",
763+
" Function evaluations: 24\n",
764+
" Gradient evaluations: 24\n",
765+
" Results: Logit\n",
766+
"===================================================================\n",
767+
"Model: Logit Pseudo R-squared: 1.000 \n",
768+
"Dependent Variable: species AIC: 8.0002 \n",
769+
"Date: 2020-09-03 13:17 BIC: 16.9941 \n",
770+
"No. Observations: 70 Log-Likelihood: -8.2981e-05\n",
771+
"Df Model: 3 LL-Null: -47.487 \n",
772+
"Df Residuals: 66 LLR p-value: 1.8711e-20 \n",
773+
"Converged: 1.0000 Scale: 1.0000 \n",
774+
"-------------------------------------------------------------------\n",
775+
" Coef. Std.Err. z P>|z| [0.025 0.975] \n",
776+
"-------------------------------------------------------------------\n",
777+
"sepallengthcm -2.6981 1485.3731 -0.0018 0.9986 -2913.9759 2908.5796\n",
778+
"sepalwidthcm -8.8019 2565.4623 -0.0034 0.9973 -5037.0157 5019.4118\n",
779+
"petallengthcm 12.9770 1215.7072 0.0107 0.9915 -2369.7654 2395.7193\n",
780+
"petalwidthcm 5.7296 2155.0836 0.0027 0.9979 -4218.1565 4229.6158\n",
781+
"===================================================================\n",
782+
"\n",
783+
"Model 2 Summary\n",
784+
"Iteration suggests how many loop model did to perform the fit\n",
785+
"Iterations : 22\n",
786+
"Pseudo R Square suggests overall effect size (ideal value is close to 1)\n",
787+
"Model 2, MacFadden Pseudo R Square : 1.0\n",
788+
"AIC compares Goodness of Fit, Lower AIC better is the Model\n",
789+
"Model 2, AIC : 8.0\n",
790+
"BIC also work same as AIC, Lower BIC better is the Model\n",
791+
"Model 2, BIC : 16.99\n"
792+
],
793+
"name": "stdout"
794+
}
795+
]
734796
},
735797
{
736798
"cell_type": "code",
737799
"metadata": {
738800
"id": "v67NCxRZGCXx",
739801
"colab_type": "code",
802+
"colab": {
803+
"base_uri": "https://localhost:8080/",
804+
"height": 341
805+
},
806+
"outputId": "f6a2cb25-02da-4349-d9a8-39fbd619ecca"
807+
},
808+
"source": [
809+
"# Model Prediction\n",
810+
"\n",
811+
"print(\"Sample Prediction of Model 2\")\n",
812+
"pred = result.predict(X_test)\n",
813+
"model_prediction = pd.DataFrame(pred.round(2),columns = ['Prediction'])\n",
814+
"model_prediction['temp'] = 'temp'\n",
815+
"model_prediction['Final_Class'] = np.where(model_prediction['Prediction'] > 0.5,1,0)\n",
816+
"print(model_prediction.head())\n",
817+
"temp = model_prediction.groupby('temp')['Final_Class'].apply(list)\n",
818+
"y_pred = temp.loc['temp']\n",
819+
"model_2_accuracy = accuracy_score(y_test,y_pred).round(2)\n",
820+
"print(\"\\nModel Performance\")\n",
821+
"print(\"Model 2, Accuracy :\",model_2_accuracy)\n",
822+
"model_2_precision = precision_score(y_test,y_pred).round(2)\n",
823+
"print(\"Model 2, Precision :\",model_2_precision)\n",
824+
"model_2_recall = recall_score(y_test,y_pred).round(2)\n",
825+
"print(\"Model 2, Recall :\",model_2_recall)\n",
826+
"model_2_fscore = f1_score(y_test,y_pred).round(2)\n",
827+
"print(\"Model 2, F1 Score :\",model_2_fscore)\n",
828+
"model_2_roc = roc_auc_score(y_test,y_pred)\n",
829+
"print(\"Model 2, AUC :\",model_2_roc)\n",
830+
"print(\"\\nConfusion Matrix, Model 2\")\n",
831+
"model_2_cm = confusion_matrix(y_test,y_pred)\n",
832+
"print(model_2_cm)"
833+
],
834+
"execution_count": 15,
835+
"outputs": [
836+
{
837+
"output_type": "stream",
838+
"text": [
839+
"Sample Prediction of Model 2\n",
840+
" Prediction temp Final_Class\n",
841+
"23 0.0 temp 0\n",
842+
"81 1.0 temp 1\n",
843+
"85 1.0 temp 1\n",
844+
"34 0.0 temp 0\n",
845+
"62 1.0 temp 1\n",
846+
"\n",
847+
"Model Performance\n",
848+
"Model 2, Accuracy : 1.0\n",
849+
"Model 2, Precision : 1.0\n",
850+
"Model 2, Recall : 1.0\n",
851+
"Model 2, F1 Score : 1.0\n",
852+
"Model 2, AUC : 1.0\n",
853+
"\n",
854+
"Confusion Matrix, Model 2\n",
855+
"[[21 0]\n",
856+
" [ 0 9]]\n"
857+
],
858+
"name": "stdout"
859+
}
860+
]
861+
},
862+
{
863+
"cell_type": "code",
864+
"metadata": {
865+
"id": "ca71zTREHpbi",
866+
"colab_type": "code",
740867
"colab": {}
741868
},
742869
"source": [
743-
""
870+
"# Model Comparision\n",
871+
"\n",
872+
"\n",
873+
"print(\"Model 1, Accuracy :\",model_1_accuracy)\n",
874+
"model_1_precision = precision_score(y_test,y_pred).round(2)\n",
875+
"print(\"Model 1, Precision :\",model_1_precision)\n",
876+
"model_1_recall = recall_score(y_test,y_pred).round(2)\n",
877+
"print(\"Model 1, Recall :\",model_1_recall)\n",
878+
"model_1_fscore = f1_score(y_test,y_pred).round(2)\n",
879+
"print(\"Model 1, F1 Score :\",model_1_fscore)\n",
880+
"model_1_roc = roc_auc_score(y_test,y_pred)\n",
881+
"print(\"Model 1, AUC :\",model_1_roc)\n",
882+
"print(\"\\nConfusion Matrix, Model 1\")\n",
883+
"model_1_cm = confusion_matrix(y_test,y_pred)\n",
884+
"print(model_1_cm)"
744885
],
745886
"execution_count": null,
746887
"outputs": []

0 commit comments

Comments
 (0)