added MultClassPipe2 that has one hot encoder step to the pipeline

MultClassPipe.py

@@ -23,6 +23,7 @@ from sklearn.model_selection import train_test_split
 from sklearn.metrics import accuracy_score, confusion_matrix, precision_score, recall_score, roc_auc_score, roc_curve, f1_score
 #%%
 rs = {'random_state': 42}
+# TODO: add preprocessing step with one hot encoder
 
 # Multiple Classification - Model Pipeline
 def MultClassPipeline(X_train, X_test, y_train, y_test):
@@ -35,6 +36,15 @@ def MultClassPipeline(X_train, X_test, y_train, y_test):
     dt = DecisionTreeClassifier(**rs)
     et = ExtraTreesClassifier(**rs)
     rf = RandomForestClassifier(**rs)
+    rf2 = RandomForestClassifier(
+                          min_samples_leaf=50,
+                          n_estimators=150,
+                          bootstrap=True,
+                          oob_score=True,
+                          n_jobs=-1,
+                          random_state=42,
+                          max_features='auto')
+    
     xgb = XGBClassifier(**rs, verbosity=0)
 
     clfs = [
@@ -46,6 +56,7 @@ def MultClassPipeline(X_train, X_test, y_train, y_test):
             ('Decision Tree', dt), 
             ('Extra Trees', et), 
             ('Random Forest', rf), 
+            ('Random Forest2', rf2), 
             ('XGBoost', xgb)
             ]
 

my_data9.py

@@ -7,7 +7,12 @@ Created on Sat Mar  5 12:57:32 2022
 """
 #%%
 # data, etc for now  comes from my_data6.py and/or my_data5.py
+#%%
+homedir = os.path.expanduser("~")
+os.chdir(homedir + "/git/ML_AI_training/")
 
+# my function
+from MultClassPipe2 import MultClassPipeline2 
 #%% try combinations
 #import sys, os
 #os.system("imports.py")
@@ -45,11 +50,19 @@ X_train, X_test, y_train, y_test = train_test_split(all_features_df,
 preprocessor = ColumnTransformer(
     transformers=[
         ('num', MinMaxScaler()  , numerical_features_names)
-        #,('cat', OneHotEncoder(), categorical_features_names)
+        ,('cat', OneHotEncoder(), categorical_features_names)
-        ])
+        ], remainder = 'passthrough')
+
+f = preprocessor.fit(numerical_features_df)
+f2 = preprocessor.transform(numerical_features_df)
+
+f3 = preprocessor.fit_transform(numerical_features_df)
+(f3==f2).all()
+
+f4 = preprocessor.fit_transform(all_features_df)
+f4
+reprocessor.fit_transform(numerical_features_df)
 
-preprocessor.fit(numerical_features_df)
-preprocessor.transform(numerical_features_df)
 #%%
 model_log = Pipeline(steps = [
     ('preprocess', preprocessor)
@@ -90,21 +103,30 @@ output = cross_validate(model, X_trainN, y_trainN
                         , cv = 10
                         , return_train_score = False)
 pd.DataFrame(output).mean()
+#%% Run multiple models using MultClassPipeline
+# only good for numerical features as categ features is not supported yet!
+t1_res = MultClassPipeline2(X_trainN, X_testN, y_trainN, y_testN, input_df = all_features_df)
+t1_res
 
-
+t2_res = MultClassPipeline2(X_train, X_test, y_train, y_test, input_df = all_features_df)
+t2_res
 #%%
+# https://machinelearningmastery.com/columntransformer-for-numerical-and-categorical-data/
+#Each transformer is a three-element tuple that defines the name of the transformer, the transform to apply, and the column indices to apply it to. For example:
+# (Name, Object, Columns)
 
-selector_logistic = RFECV(estimator = model
+# Determine categorical and numerical features
-                       , cv = 10
+numerical_ix = all_features_df.select_dtypes(include=['int64', 'float64']).columns
-                       , step = 1)
+numerical_ix
+categorical_ix = all_features_df.select_dtypes(include=['object', 'bool']).columns
+categorical_ix
 
-X_trainN, X_testN, y_trainN, y_testN = train_test_split(numerical_features_df 
+# Define the data preparation for the columns
-                                                    , target1 
+t = [('cat', OneHotEncoder(), categorical_ix)
-                                                    , test_size = 0.33
+     , ('num', MinMaxScaler(), numerical_ix)]
-                                                    , random_state = 42)
+col_transform = ColumnTransformer(transformers=t
+                                  , remainder='passthrough')
+# create pipeline (unlike example above where the col transfer was a preprocess step and it was fit_transformed)             
 
-selector_logistic_xtrain = selector_logistic.fit_transform(X_trainN, y_trainN)
+pipeline = Pipeline(steps=[('prep', col_transform)
-print(sel_rfe_logistic.get_support())
+                                   , ('classifier', clf)])
-X_trainN.columns
-
-print(sel_rfe_logistic.ranking_)

p_jr_d1.py

@@ -351,6 +351,7 @@ pred
 
 # make a pipeline
 # PCA(Dimension reduction to two) -> Scaling the data -> DecisionTreeClassification
+#https://www.geeksforgeeks.org/pipelines-python-and-scikit-learn/
 pipe1 = Pipeline([('pca', PCA(n_components = 2))
                  , ('std', StandardScaler())
                  , ('decision_tree', DecisionTreeClassifier())]