Comparing classifiers

evaluating_classifiers.py

# this is a custom module
import assess_clf_models as acm

from sklearn.linear_model import LogisticRegression
from sklearn.tree import DecisionTreeClassifier
from sklearn.ensemble import GradientBoostingClassifier, RandomForestClassifier, BaggingClassifier
from catboost import CatBoostClassifier

from sklearn.compose import ColumnTransformer
from sklearn.preprocessing import StandardScaler, OneHotEncoder, FunctionTransformer
from sklearn.pipeline import Pipeline, FeatureUnion
from sklearn.model_selection import train_test_split
from sklearn.feature_extraction.text import CountVectorizer

# define a preprocessing pipeline
  # see https://github.com/kbfreder/small-molecules/blob/master/02c-Pipeline_ModelEval_Imbalance.ipynb
  # for an example
pipe_with_scale = Pipeline([
    ('all', FeatureUnion([
        ('cvs', preproc_cv),
        ('feat_eng', preproc_feat_eng_scaled),
        ('ohe', preproc_ohe), 
        ('pass', preproc_pass),
        ('num', preproc_scale),
    ])
    )
])

log_reg = LogisticRegression(solver='liblinear')
rand_for = RandomForestClassifier(n_estimators=10)
grad_boost = GradientBoostingClassifier()
cb = CatBoostClassifier(cat_features=cb_cat_cols, logging_level='Silent')

mini_results_lr = acm.assess_model(pipe_with_scale, log_reg, mini_X, mini_y)
mini_results_rf = acm.assess_model(pipe_cat_only, rand_for, mini_X, mini_y)
mini_results_gb = acm.assess_model(pipe_cat_only, grad_boost, mini_X, mini_y)
mini_results_cb = acm.assess_model(pipe_cat_boost, cb, mini_X, mini_y)

comp_df = pd.DataFrame([mini_results_lr, mini_results_rf, mini_results_gb, mini_results_cb],
             index=['LogReg','RandFor','GradBst', 'Cat Boost'])