Incorrect total time in Sklearn GridSearchCV

Question

I'm trying to use sklearn GridSearchCV to find the best parameters for an Elastic Net Classifier. However, I'm worried something is wrong since the total time stamp in the logs keeps resetting, showing very low values like 0.5s, 0.7s...

I'm running the following code:

from sklearn.model_selection import LeaveOneGroupOut
from sklearn.model_selection import GridSearchCV
from sklearn.linear_model import SGDClassifier
import pandas as pd
from sklearn.utils import shuffle

path = "example.csv"

seed = 27

df = pd.read_csv(path)
df = shuffle(df)

X = df.drop(columns=["target", "group"])
y = df.target
groups = df.group

# Define parameters to evaluate
alpha_values = [0.0001, 0.00025, 0.0005, 0.00075, 0.001, 0.005, 0.0025, 0.0075, 0.01, 0.05, 0.1, 1]
n_iter_no_change_values = [20, 50, 100, 150]
l1_ratio_values = [0.049, 0.10, 0.15, 0.20, 0.25, 0.30, 0.35, 0.40, 0.45, 0.50,
                   0.55, 0.60, 0.65, 0.70, 0.75, 0.80, 0.85, 0.89, 0.95, 0.99]

parameters = {'penalty': ['elasticnet'],
              'class_weight': ['balanced'],
              'loss': ['log'],
              'random_state': [seed],
              'alpha': alpha_values,
              'l1_ratio': l1_ratio_values,
              'n_iter_no_change': n_iter_no_change_values,
              'max_iter': [2500]
              }

elasticnet = SGDClassifier()
logo = LeaveOneGroupOut()

clf = GridSearchCV(estimator=elasticnet,
                   param_grid=parameters,
                   scoring='roc_auc_ovo_weighted',
                   cv=logo.split(X=X, groups=groups),
                   verbose=3,
                   n_jobs=-1,
                   pre_dispatch='n_jobs'
                   )

clf.fit(X, y)

The example file needed to to run the code is downloadable here.

Producing the following log:

Fitting 2 folds for each of 960 candidates, totalling 1920 fits
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.049, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.578 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.049, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.615 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.049, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.611 total time=   0.4s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.049, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.579 total time=   0.4s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.1, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.610 total time=   0.2s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.1, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.568 total time=   0.2s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.049, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.608 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.049, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.580 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.1, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.574 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.1, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.609 total time=   0.4s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.049, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.605 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.049, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.580 total time=   0.6s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.1, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.606 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.1, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.577 total time=   0.5s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.15, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.610 total time=   0.2s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.15, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.568 total time=   0.2s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.15, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.573 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.15, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.611 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.1, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.604 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.1, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.578 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.2, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.575 total time=   0.2s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.15, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.608 total time=   0.5s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.2, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.621 total time=   0.2s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.15, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.578 total time=   0.5s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.2, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.615 total time=   0.4s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.15, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.606 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.15, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.580 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.2, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.575 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.25, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.611 total time=   0.2s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.2, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.580 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.25, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.572 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.2, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.609 total time=   0.5s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.25, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.610 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.2, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.608 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.25, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.574 total time=   0.4s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.2, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.580 total time=   0.7s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.3, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.572 total time=   0.2s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.3, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.619 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.25, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.606 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.25, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.579 total time=   0.5s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.3, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.618 total time=   0.3s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.3, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.571 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.25, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.605 total time=   0.6s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.35, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.611 total time=   0.3s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.25, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.579 total time=   0.7s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.3, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.613 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.3, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.572 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.35, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.565 total time=   0.2s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.35, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.611 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.3, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.609 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.3, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.573 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.35, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.567 total time=   0.3s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.35, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.567 total time=   0.5s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.35, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.610 total time=   0.5s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.4, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.620 total time=   0.3s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.4, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.564 total time=   0.2s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.4, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.565 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.4, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.620 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.35, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.606 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.35, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.569 total time=   0.6s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.45, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.619 total time=   0.2s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.45, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.565 total time=   0.2s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.45, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.621 total time=   0.3s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.4, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.568 total time=   0.5s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.4, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.613 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.45, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.568 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.4, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.610 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.4, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.570 total time=   0.7s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.5, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.571 total time=   0.2s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.5, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.616 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.45, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.619 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.45, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.570 total time=   0.4s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.5, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.618 total time=   0.4s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.5, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.572 total time=   0.3s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.45, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.571 total time=   0.6s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.45, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.616 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.5, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.570 total time=   0.4s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.5, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.617 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.55, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.567 total time=   0.2s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.55, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.619 total time=   0.3s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.55, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.568 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.5, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.616 total time=   0.6s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.55, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.622 total time=   0.4s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.5, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.570 total time=   0.6s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.6, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.616 total time=   0.2s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.6, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.566 total time=   0.3s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.55, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.569 total time=   0.5s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.6, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.619 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.55, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.622 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.6, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.568 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.55, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.621 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.55, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.570 total time=   0.6s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.65, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.612 total time=   0.2s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.65, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.572 total time=   0.2s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.6, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.622 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.6, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.566 total time=   0.4s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.65, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.617 total time=   0.3s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.65, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.574 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.6, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.623 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.6, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.567 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.65, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.574 total time=   0.5s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.7, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.605 total time=   0.2s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.65, loss=log, max_iter=2500, n_iter_no_change=100, penalty=elasticnet, random_state=27;, score=0.617 total time=   0.5s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.7, loss=log, max_iter=2500, n_iter_no_change=20, penalty=elasticnet, random_state=27;, score=0.573 total time=   0.2s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.7, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.607 total time=   0.3s
[CV 1/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.65, loss=log, max_iter=2500, n_iter_no_change=150, penalty=elasticnet, random_state=27;, score=0.619 total time=   0.6s
[CV 2/2] END alpha=0.0001, class_weight=balanced, l1_ratio=0.7, loss=log, max_iter=2500, n_iter_no_change=50, penalty=elasticnet, random_state=27;, score=0.573 total time=   0.3s

As you can see the total time does not increase at it should, rather it seems to reset every line. Is this due to a bug in sklearn or is there something wrong in my execution?

user_13970089 · Answer

total_time means fit_time + score_time for each model trained in the grid search CV that's why it is not increasing. One can also see that it is almost the same time for the two parts of the CV with the same hyperparameters. If you want to have a look in the code : https://github.com/scikit-learn/scikit-learn/blob/17df37aee774720212c27dbc34e6f1feef0e2482/sklearn/model_selection/_validation.py#L714

Incorrect total time in Sklearn GridSearchCV

1 Answers