ProductTable.append_column('Difference From Fresh20', 20) pandasDF_withColumnDifference_From_Fresh20 = ProductTable.to_df() X_New_Prediction = pandasDF_withColumnDifference_From_Fresh20[[ 'ProductTypeA', 'ProductTypeB', 'ProductTypeC', 'ProductTypeD', 'ProductTypeE', 'ProductTypeF', 'ProductTypeG', 'ProductTypeH', 'ProductTypeI', 'BaseIngredientA', 'BaseIngredientB', 'BaseIngredientC', 'BaseIngredientD', 'BaseIngredientE', 'BaseIngredientF', 'ProcessTypeA', 'ProcessTypeB', 'ProcessTypeC', 'Difference From Fresh20', 'WarmClimate', 'ColdClimate', 'HighTemperatureandHumidity', 'PackagingStabilizerAdded', 'PackagingStabilizerNotAdded', 'Processing Agent Stability Index', 'PreservativeAdded', 'PreservativeNotAdded' ]] #NEW COLOMN ShelfLife AS PREDICTED OUTPUT FROM MODEL ProductTable.append_column('ShelfLifeInWeeks', clf1.predict(X_New_Prediction)) ProductTable = ProductTable.move_column('ShelfLifeInWeeks', 2) ProductTableOut.append_column('Prediction', clf1.predict(X_New_Prediction)) ProductTableOut['Prediction'] = ProductTableOut.apply(lambda x: math.floor(x), 'Prediction') ProductTableOut.show(5) #RELATIVE IMPORTANCE OF VARIABLES IN THE MODEL print("RELATIVE IMPORTANCE OF VARIABLES IN THE MODEL") importances = pd.DataFrame({ 'feature': X_Train_DTC.columns, 'importance': np.round(clf1.feature_importances_, 3) }) importances = importances.sort_values('importance', ascending=False) print(importances)
def move_column(self, *args, **kwargs): return self._fix_(Table.move_column(self, *args, **kwargs))