Upload events_classification_biotech.py

events_classification_biotech.py

@@ -1,5 +1,5 @@
-import pandas as pd
 import datasets
+from datasets import load_dataset
 from datasets.tasks import TextClassification
 
 DESCRIPTION = """
@@ -75,25 +75,15 @@ class BiotechNews(datasets.GeneratorBasedBuilder):
             datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={"filepath": test_path}),
         ]
 
-    def convert_to_none(self, label):
-        if type(label) == float:
-              label = None
-        return label
-
     def get_all_labels(self, curr_labels):
          curr_labels = [lbl for lbl in curr_labels if lbl]
          all_curr_labels = ', '.join(curr_labels)
          return all_curr_labels
     
     def _generate_examples(self, filepath):
-            csv_reader = pd.read_csv(filepath)
-            for id_, row in csv_reader.iterrows():
-                i, title, content, organization, label1, label2, label3, label4, label5 = row
-                label1 = self.convert_to_none(label1)
-                label2 = self.convert_to_none(label2)
-                label3 = self.convert_to_none(label3)
-                label4 = self.convert_to_none(label4)
-                label5 = self.convert_to_none(label5)
+            csv_reader = load_dataset('csv',data_files=filepath, split='train')
+            for example in csv_reader:
+                id_, title, content, organization, label1, label2, label3, label4, label5 = example.values()
 
                 curr_labels = (label1, label2, label3, label4, label5)
                 all_curr_labels = self.get_all_labels(curr_labels)