Projeto #3: Breast cancer com tunning, gernado um IndexError

[tvaditya]

Olá

 

Estou encontrando um IndexError mesmo tendo usado as versões exatas da aula, que seria esse:

/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py:774: UserWarning: Scoring failed. The score on this train-test partition for these parameters will be set to nan. Details:
Traceback (most recent call last):
File “/usr/local/lib/python3.7/dist-packages/sklearn/model_selection/_validation.py”, line 761, in _score
scores = scorer(estimator, X_test, y_test)
File “/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_scorer.py”, line 221, in __call__
sample_weight=sample_weight,
File “/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_scorer.py”, line 258, in _score
y_pred = method_caller(estimator, “predict”, X)
File “/usr/local/lib/python3.7/dist-packages/sklearn/metrics/_scorer.py”, line 68, in _cached_call
return getattr(estimator, method)(*args, **kwargs)
File “/usr/local/lib/python3.7/dist-packages/skorch/classifier.py”, line 357, in predict
return (y_proba[:, 1] > self.threshold).astype(‘uint8’)
IndexError: too many indices for array: array is 1-dimensional, but 2 were indexed

Gostaria de entender a razão, usei a versão para pytorch 1.4.0 e o skorch 0.7.0.  ,  ele aponta erro  na hora de fazer o fit do GridSearch

grid_search.fit(previsores, classe)

[Denny Ceccon]

Olá,

Este erro é relacionado à última versão do Skorch, agora o arquivo Colab já está atualizado, obrigado por nos avisar.

[Autor]

Posts