pycaret compare_models() call doesn't recognize sort

Question

After creating the clr_default:

clr_default = setup(df_rain_definitivo_one_drop_catboost_norm_fs_dropna,fold_shuffle=True, target='RainTomorrow', session_id=123)

I've tried to use the compare_models() function in Pycaret, using the following call:

best_model = compare_models()
from pycaret.classification import *

However I get the following error message:

ValueError                                Traceback (most recent call last)
<ipython-input-228-e1d76b68915a> in <module>()
----> 1 best_model = compare_models(n_select = 5, sort='Accuracy')

1 frames
/usr/local/lib/python3.7/dist-packages/pycaret/internal/tabular.py in compare_models(include, exclude, fold, round, cross_validation, sort, n_select, budget_time, turbo, errors, fit_kwargs, groups, verbose, display)
   1954         if sort is None:
   1955             raise ValueError(
-> 1956                 f"Sort method not supported. See docstring for list of available parameters."
   1957             )
   1958 

ValueError: Sort method not supported. See docstring for list of available parameters.

I've tried to call compare_models() with the sort parameter = 'Accuracy' but it didn't do any good. Also, I'm on Google Colab

score 0 · Answer 1 · edited May 22 '22 at 11:27

I dont get what is n_select = 5? do you want to get the top-5 models? Otherwise;

Im using your code examples: First import pycaret

from pycaret.classification import *

Then setup,

clr_default = setup(df_rain_definitivo_one_drop_catboost_norm_fs_dropna,fold_shuffle=True, target='RainTomorrow', session_id=123)

Last use compare model method

best_model = compare_models(sort='Accuracy')

After that you can create your models then tune it.

pycaret compare_models() call doesn't recognize sort

1 Answers1