feat/336 : check if pandas df and drop_last default to True

pytorch_tabnet/abstract_model.py

@@ -13,6 +13,7 @@
     create_dataloaders,
     define_device,
     ComplexEncoder,
+    check_input
 )
 from pytorch_tabnet.callbacks import (
     CallbackContainer,
@@ -22,7 +23,7 @@
 )
 from pytorch_tabnet.metrics import MetricContainer, check_metrics
 from sklearn.base import BaseEstimator
-from sklearn.utils import check_array
+
 from torch.utils.data import DataLoader
 import io
 import json
@@ -115,7 +116,7 @@ def fit(
         batch_size=1024,
         virtual_batch_size=128,
         num_workers=0,
-        drop_last=False,
+        drop_last=True,
         callbacks=None,
         pin_memory=True,
         from_unsupervised=None,
@@ -182,7 +183,7 @@ def fit(
         else:
             self.loss_fn = loss_fn
 
-        check_array(X_train)
+        check_input(X_train)
 
         self.update_fit_params(
             X_train,

pytorch_tabnet/pretraining.py

@@ -1,12 +1,12 @@
 import torch
 import numpy as np
-from sklearn.utils import check_array
 from torch.utils.data import DataLoader
 from pytorch_tabnet import tab_network
 from pytorch_tabnet.utils import (
     create_explain_matrix,
     filter_weights,
-    PredictDataset
+    PredictDataset,
+    check_input
 )
 from torch.nn.utils import clip_grad_norm_
 from pytorch_tabnet.pretraining_utils import (
@@ -55,7 +55,7 @@ def fit(
         batch_size=1024,
         virtual_batch_size=128,
         num_workers=0,
-        drop_last=False,
+        drop_last=True,
         callbacks=None,
         pin_memory=True,
     ):
@@ -118,7 +118,7 @@ def fit(
         else:
             self.loss_fn = loss_fn
 
-        check_array(X_train)
+        check_input(X_train)
 
         self.update_fit_params(
             weights,

pytorch_tabnet/pretraining_utils.py

@@ -2,8 +2,8 @@
 from pytorch_tabnet.utils import (
     create_sampler,
     PredictDataset,
+    check_input
 )
-from sklearn.utils import check_array
 
 
 def create_dataloaders(
@@ -93,7 +93,7 @@ def validate_eval_set(eval_set, eval_name, X_train):
     ), "eval_set and eval_name have not the same length"
 
     for set_nb, X in enumerate(eval_set):
-        check_array(X)
+        check_input(X)
         msg = (
             f"Number of columns is different between eval set {set_nb}"
             + f"({X.shape[1]}) and X_train ({X_train.shape[1]})"

pytorch_tabnet/utils.py

@@ -5,6 +5,7 @@
 import scipy
 import json
 from sklearn.utils import check_array
+import pandas as pd
 
 
 class TorchDataset(Dataset):
@@ -271,7 +272,7 @@ 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_array(X)
+        check_input(X)
         msg = (
             f"Dimension mismatch between X_{name} "
             + f"{X.shape} and X_train {X_train.shape}"
@@ -337,3 +338,15 @@ def default(self, obj):
             return int(obj)
         # Let the base class default method raise the TypeError
         return json.JSONEncoder.default(self, obj)
+
+
+def check_input(X):
+    """
+    Raise a clear error if X is a pandas dataframe
+    and check array according to scikit rules
+    """
+    if isinstance(X, (pd.DataFrame, pd.Series)):
+        err_message = "Pandas DataFrame are not supported: apply X.values when calling fit"
+        raise(ValueError, err_message)
+    check_array(X)
+    return