fix: top4-balanced

README.md

@@ -4,18 +4,8 @@ language:
 task_categories:
 - text-classification
 
-configs:
-- config_name: original
-  data_files:
-  - split: complete
-    path: "data/original.csv"
-  default: true
-- config_name: top4-balanced
-  data_files:
-  - split: complete
-    path: "data/top4-balanced.csv"
-
 dataset_info:
+- config_name: original
   features:
   - name: source
     dtype: string
@@ -53,4 +43,44 @@ dataset_info:
           '11': Top Stories
           '12': U.S.
           '13': World
+- config_name: top4-balanced
+  features:
+  - name: source
+    dtype: string
+  - name: url
+    dtype: string
+  - name: title
+    dtype: string
+  - name: image
+    dtype: string
+  - name: category
+    dtype: string
+  - name: description
+    dtype: string
+  - name: rank
+    dtype: int32
+  - name: pubdate
+    dtype: string
+  - name: text
+    dtype: string
+  - name: label
+    dtype:
+      class_label:
+        names:
+          '0': World
+          '1': Sports
+          '2': Business
+          '3': Sci/Tech
+
+configs:
+- config_name: original
+  data_files:
+  - split: complete
+    path: "data/original.csv"
+  default: true
+- config_name: top4-balanced
+  data_files:
+  - split: complete
+    path: "data/top4-balanced.csv"
+
 ---

convert.py

@@ -29,6 +29,7 @@ HEADER = [
 
 OUTPUT_FILE_PATH = os.path.join("data", "original.csv")
 
+
 def _clean_text(text):
     text = text.replace("\\\n", "\n")
     text = html.unescape(text)
@@ -58,9 +59,9 @@ def _clean_html(text):
 
         if output == "null":
             return ""
-        
+
         return output
-    
+
     return ""
 
 
@@ -125,7 +126,9 @@ def run():
         row["label"] = _categories.index(row["category"])
 
     save_csv(rows)
-    save_csv_categories(["World", "Sports", "Business", "Sci/Tech"], "top4-balanced", is_balanced=True)
+    save_csv_categories(
+        ["World", "Sports", "Business", "Sci/Tech"], "top4-balanced", is_balanced=True
+    )
 
 
 def save_csv(rows, fname=OUTPUT_FILE_PATH):
@@ -133,7 +136,7 @@ def save_csv(rows, fname=OUTPUT_FILE_PATH):
     Save the processed data into a CSV file.
     """
     os.makedirs("data", exist_ok=True)
-    
+
     with open(fname, "w", encoding="utf8") as f:
         writer = csv.DictWriter(f, fieldnames=rows[0].keys())
         writer.writeheader()
@@ -163,6 +166,10 @@ def save_csv_categories(categories, config_name, is_balanced=True, **kwargs):
     else:
         df = df[df["category"].isin(categories)]
 
+    # remap the labels to 0, 1, 2, ...
+    for i, row in df.iterrows():
+        df.at[i, "label"] = categories.index(row["category"])
+
     df.to_csv(os.path.join("data", f"{config_name}.csv"), index=False)
 
 

data/top4-balanced.csv

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d877f38b2e29c8cbf7d7334f846158c7eb7c463ad5e51827d41dfb54b3006dd2
-size 343796732
+oid sha256:3c900ade9ea2c83255b09021e70d80d58b933880655fc6e794e12a652a30d33f
+size 343919424