[FIXED] Attribute Error: Pipeline object has no attribute 'best_estimator_'

Issue

Could someone please demonstrate how to amend this code (which is a reproducible example):

from sklearn.model_selection import GridSearchCV 
from sklearn.model_selection import StratifiedKFold,KFold
from sklearn.feature_selection import SelectKBest 
#from xgboost import XGBClassifier 
from sklearn.feature_selection import mutual_info_classif 
from sklearn.feature_selection import SelectKBest, RFECV 
from sklearn.pipeline import Pipeline 
from sklearn.svm import SVC 
from sklearn.ensemble import RandomForestClassifier 
from sklearn.naive_bayes import GaussianNB 
from sklearn.linear_model import LogisticRegression 
from sklearn.metrics import make_scorer, recall_score, accuracy_score, precision_score 
from sklearn.model_selection import StratifiedKFold 
from sklearn.metrics import make_scorer 
from sklearn.metrics import precision_score,recall_score,f1_score,roc_auc_score 
from sklearn import metrics
from sklearn.datasets import make_classification
from numpy import mean 
from sklearn.model_selection import train_test_split
from numpy import std 
from sklearn.utils import shuffle 
import numpy as np 
from sklearn.metrics import roc_curve
from sklearn.pipeline import Pipeline
import matplotlib.pyplot as plt 
import pandas as pd



full_X_train,full_y_train = make_classification(n_samples =500,n_features = 20, random_state=1, n_informative=10,n_redundant=10)

def run_model_with_grid_search(param_grid={},output_plt_file = 'plt.png',model_name=RandomForestClassifier(),X_train=full_X_train,y_train=full_y_train,model_id='random_forest_with_hpo_no_fs_geno_class', n_splits=5, output_file='random_forest_with_hpo_no_fs_geno_class.txt'): 
  

      cv_outer = KFold(n_splits=5,shuffle=True,random_state=1)

      for train_ix,test_ix in cv_outer.split(X_train):
              split_x_train, split_x_test = X_train[train_ix,:],X_train[test_ix,:] #add in .iloc               
              split_y_train, split_y_test = y_train[train_ix],y_train[test_ix]  #add in .iloc

              cv_inner = KFold(n_splits=3,shuffle=True,random_state=1)
              model = model_name
              rfecv = RFECV(estimator=model, step=1,cv=5, scoring='roc_auc')
              search = GridSearchCV(model,param_grid=param_grid,scoring='roc_auc',cv=cv_inner,refit=True)
              pipeline = Pipeline([('feature_sele',rfecv),('clf_cv',search)])
              result = pipeline.fit(split_x_train,split_y_train)
              #result = search.fit(split_x_train,split_y_train)
              best_model = result.best_estimator_
              yhat = best_model.predict(split_x_test)

              print('>acc=%.3f,est=%.3f,cfg=%s'%(accuracy,result.best_score_,result.best_params_))

      return


param_grid = [{
               'min_samples_leaf':[1,3,5],
              }]

run_model_with_grid_search(param_grid=param_grid)
    

Generates:

Attribute Error: Pipeline object has no attribute 'best_estimator_'

The ultimate aim is to perform nested cross validation, hyper parameter optimization and feature selection in this function, and I was trying to follow this example, if someone could show me how to edit this function to perform that correctly I'd appreciate it.

Solution

Normally, you'd run grid search on the pipeline, not the pipeline on grid search. Is there a certain reason you'd want it the other way round?

pipeline = Pipeline([('feature_sele',rfecv), ('clf',model)])
search = GridSearchCV(pipeline, param_grid=param_grid, scoring='roc_auc', cv=cv_inner, refit=True)
result = search.fit(split_x_train, split_y_train)
best_model = result.best_estimator_

(param_grid will require clf_ prefix to hyperparameter names ofc.)

On an unlrelated note, accuracy seems to be undefined.

Answered By - dx2-66