3

I run Boruta with RandomForestClassifier the previous day on my data (nb features = 36) and got 17/36 confirmed. Now I run it again and there is 0/36 and stop at the 9th iteration. Any idea why this is happening? 

%%time
rfc = RandomForestClassifier(n_estimators=200, n_jobs=-1, class_weight= 'balanced_subsample' , max_depth=10)

boruta_selector = BorutaPy(rfc, n_estimators='auto', verbose=2, random_state=1)

boruta_selector_before.fit(X, y)

the current output look is the follow

Iteration:  1 / 100
Confirmed:  0
Tentative:  36
Rejected:   0
Iteration:  2 / 100
Confirmed:  0
Tentative:  36
Rejected:   0
Iteration:  3 / 100
Confirmed:  0
Tentative:  36
Rejected:   0
Iteration:  4 / 100
Confirmed:  0
Tentative:  36
Rejected:   0
Iteration:  5 / 100
Confirmed:  0
Tentative:  36
Rejected:   0
Iteration:  6 / 100
Confirmed:  0
Tentative:  36
Rejected:   0
Iteration:  7 / 100
Confirmed:  0
Tentative:  36
Rejected:   0
Iteration:  8 / 100
Confirmed:  0
Tentative:  0
Rejected:   36


BorutaPy finished running.

Iteration:  9 / 100
Confirmed:  0
Tentative:  0
Rejected:   36
CPU times: user 1.95 s, sys: 377 ms, total: 2.33 s
Wall time: 2.79 s
irkinosor
  • 233
  • 1
  • 7

1 Answers1

1

My mistake came from the fact that my target variable sample only had one class. To reduce computation time, I took a sample from the main data.

irkinosor
  • 233
  • 1
  • 7