changes for test_dataloading warnings

amaiya · Jun 14, 2024 · 6a3a649 · 6a3a649
1 parent 2b1a580
commit 6a3a649
Show file tree

Hide file tree

Showing 3 changed files with 9 additions and 7 deletions.
diff --git a/ktrain/text/data.py b/ktrain/text/data.py
@@ -292,10 +292,10 @@ def texts_from_df(
 
     # read in train and test data
     train_df = train_df.copy()
-    train_df[text_column].fillna("fillna", inplace=True)
+    train_df[text_column] = train_df[text_column].fillna("fillna")
     if val_df is not None:
         val_df = val_df.copy()
-        val_df[text_column].fillna("fillna", inplace=True)
+        val_df[text_column] = val_df[text_column].fillna("fillna")
     else:
         train_df, val_df = train_test_split(
             train_df, test_size=val_pct, random_state=random_state

diff --git a/ktrain/text/ner/preprocessor.py b/ktrain/text/ner/preprocessor.py
@@ -178,13 +178,13 @@ def conll2003_to_df(filepath, encoding="latin1"):
                 if not docstart:
                     sent_id += 1
     df = pd.DataFrame({SENT_COL: sents, WORD_COL: words, TAG_COL: tags})
-    df = df.fillna(method="ffill")
+    df = df.ffill()
     return df
 
 
 def gmb_to_df(filepath, encoding="latin1"):
     df = pd.read_csv(filepath, encoding=encoding)
-    df = df.fillna(method="ffill")
+    df = df.ffill()
     return df
 
 
@@ -234,7 +234,9 @@ def __init__(self, data, word_column, tag_column, sentence_column):
                 s[word_column].values.tolist(), s[tag_column].values.tolist()
             )
         ]
-        self.grouped = self.data.groupby(sentence_column).apply(agg_func)
+        self.grouped = self.data.groupby(sentence_column).apply(
+            agg_func, include_groups=False
+        )
         self.sentences = [s for s in self.grouped]
 
     def get_next(self):

diff --git a/ktrain/vision/data.py b/ktrain/vision/data.py
@@ -597,7 +597,7 @@ def images_from_df(
         x_col=image_column,
         y_col=label_columns,
         target_size=target_size,
-        class_mode="other",
+        class_mode="raw",
         shuffle=True,
         interpolation="bicubic",
         color_mode=color_mode,
@@ -611,7 +611,7 @@ def images_from_df(
             x_col=image_column,
             y_col=label_columns,
             target_size=target_size,
-            class_mode="other",
+            class_mode="raw",
             shuffle=False,
             interpolation="bicubic",
             color_mode=color_mode,