feat: add check nan and inf

pytorch_tabnet/abstract_model.py

@@ -11,6 +11,7 @@
  create_explain_matrix,
  validate_eval_set,
  create_dataloaders,
+ check_nans,
 )
 from pytorch_tabnet.callbacks import (
  CallbackContainer,
@@ -140,6 +141,8 @@ def fit(
  else:
  self.loss_fn = loss_fn
 
+ check_nans(X_train)
+ check_nans(y_train)
  self.update_fit_params(
  X_train, y_train, eval_set, weights,
  )

pytorch_tabnet/utils.py

@@ -237,6 +237,8 @@ def validate_eval_set(eval_set, eval_name, X_train, y_train):
  len(elem) == 2 for elem in eval_set
  ), "Each tuple of eval_set need to have two elements"
  for name, (X, y) in zip(eval_name, eval_set):
+ check_nans(X)
+ check_nans(y)
  msg = (
  f"Number of columns is different between X_{name} "
  + f"({X.shape[1]}) and X_train ({X_train.shape[1]})"
@@ -255,3 +257,10 @@ def validate_eval_set(eval_set, eval_name, X_train, y_train):
  assert X.shape[0] == y.shape[0], msg
 
  return eval_name, eval_set
+
+
+def check_nans(array):
+ if np.isnan(array).any():
+ raise ValueError("NaN were found, TabNet does not allow nans.")
+ if np.isinf(array).any():
+ raise ValueError("Infinite values were found, TabNet does not allow inf.")