igorcs commited on
Commit
3c3e56c
·
verified ·
1 Parent(s): 5f2702b

Adding general and specific comments to Source A

Browse files
Files changed (1) hide show
  1. aes_enem_dataset.py +6 -2
aes_enem_dataset.py CHANGED
@@ -142,6 +142,8 @@ class AesEnemDataset(datasets.GeneratorBasedBuilder):
142
  "essay_text": datasets.Value("string"),
143
  "grades": datasets.Sequence(datasets.Value("int16")),
144
  "essay_year": datasets.Value("int16"),
 
 
145
  }
146
  )
147
 
@@ -286,7 +288,6 @@ class AesEnemDataset(datasets.GeneratorBasedBuilder):
286
 
287
  grader_a = pd.read_csv(f"{dirname}/GraderA.csv")
288
  grader_b = pd.read_csv(f"{dirname}/GraderB.csv")
289
-
290
  for grader in [grader_a, grader_b]:
291
  grader.grades = grader.grades.apply(lambda x: x.strip("[]").split(", "))
292
  grader.grades = grader.grades.apply(map_list)
@@ -393,6 +394,7 @@ class AesEnemDataset(datasets.GeneratorBasedBuilder):
393
  assert (
394
  len(set(val_df["id_prompt"]).intersection(set(test_df["id_prompt"]))) == 0
395
  ), "Overlap between val and test id_prompt"
 
396
  train_df.to_csv(f"{dirname}/train.csv", index=False)
397
  val_df.to_csv(f"{dirname}/validation.csv", index=False)
398
  test_df.to_csv(f"{dirname}/test.csv", index=False)
@@ -415,6 +417,8 @@ class AesEnemDataset(datasets.GeneratorBasedBuilder):
415
  "essay_text": row["essay"],
416
  "grades": grades,
417
  "essay_year": row["essay_year"],
 
 
418
  }
419
 
420
 
@@ -585,7 +589,7 @@ class HTMLParser:
585
  for span in soup.find_all("span"):
586
  span.decompose()
587
  result = table.find_all("p")
588
- result = "\n".join(
589
  [paragraph.get_text().strip() for paragraph in result]
590
  )
591
  return result
 
142
  "essay_text": datasets.Value("string"),
143
  "grades": datasets.Sequence(datasets.Value("int16")),
144
  "essay_year": datasets.Value("int16"),
145
+ "general_comment": datasets.Value("string"),
146
+ "specific_comment": datasets.Value("string"),
147
  }
148
  )
149
 
 
288
 
289
  grader_a = pd.read_csv(f"{dirname}/GraderA.csv")
290
  grader_b = pd.read_csv(f"{dirname}/GraderB.csv")
 
291
  for grader in [grader_a, grader_b]:
292
  grader.grades = grader.grades.apply(lambda x: x.strip("[]").split(", "))
293
  grader.grades = grader.grades.apply(map_list)
 
394
  assert (
395
  len(set(val_df["id_prompt"]).intersection(set(test_df["id_prompt"]))) == 0
396
  ), "Overlap between val and test id_prompt"
397
+ #train_df['essay_year'] = train_df['essay_year'].astype(int)
398
  train_df.to_csv(f"{dirname}/train.csv", index=False)
399
  val_df.to_csv(f"{dirname}/validation.csv", index=False)
400
  test_df.to_csv(f"{dirname}/test.csv", index=False)
 
417
  "essay_text": row["essay"],
418
  "grades": grades,
419
  "essay_year": row["essay_year"],
420
+ "general_comment": row["general"],
421
+ "specific_comment": row["specific"],
422
  }
423
 
424
 
 
589
  for span in soup.find_all("span"):
590
  span.decompose()
591
  result = table.find_all("p")
592
+ result = " ".join(
593
  [paragraph.get_text().strip() for paragraph in result]
594
  )
595
  return result