shaffei commited on
Commit
b03b53b
·
verified ·
1 Parent(s): 5d2b646

fix random splits names

Browse files
Files changed (1) hide show
  1. ppb_affinity.py +7 -12
ppb_affinity.py CHANGED
@@ -39,23 +39,19 @@ class ppb_affinity(datasets.GeneratorBasedBuilder):
39
  ]
40
  elif self.config.name == "filtered_random":
41
  filepath = dl_manager.download_and_extract("filtered.csv")
42
- # Read all rows to determine splits
43
  with open(filepath, encoding="utf-8") as f:
44
  reader = csv.DictReader(f)
45
  rows = list(reader)
46
  n_total = len(rows)
47
- # Generate shuffled indices with fixed seed
48
  indices = list(range(n_total))
49
- rng = random.Random(42) # Fixed seed for reproducibility
50
  rng.shuffle(indices)
51
- # Calculate split sizes
52
  n_train = int(0.8 * n_total)
53
  n_val = int(0.1 * n_total)
54
- n_test = n_total - n_train - n_val # Handle remainder
55
- # Split indices into ranges
56
  return [
57
  datasets.SplitGenerator(
58
- name=datasets.NamedSplit("train_rand"),
59
  gen_kwargs={
60
  "filepath": filepath,
61
  "shuffled_indices": indices,
@@ -64,7 +60,7 @@ class ppb_affinity(datasets.GeneratorBasedBuilder):
64
  },
65
  ),
66
  datasets.SplitGenerator(
67
- name=datasets.NamedSplit("validation_rand"),
68
  gen_kwargs={
69
  "filepath": filepath,
70
  "shuffled_indices": indices,
@@ -73,7 +69,7 @@ class ppb_affinity(datasets.GeneratorBasedBuilder):
73
  },
74
  ),
75
  datasets.SplitGenerator(
76
- name=datasets.NamedSplit("test_rand"),
77
  gen_kwargs={
78
  "filepath": filepath,
79
  "shuffled_indices": indices,
@@ -96,9 +92,8 @@ class ppb_affinity(datasets.GeneratorBasedBuilder):
96
  del row["split"]
97
  yield idx, row
98
  elif self.config.name == "filtered_random":
99
- # Iterate over the range [split_start, split_end) in shuffled_indices
100
  for global_idx in range(split_start, split_end):
101
  original_idx = shuffled_indices[global_idx]
102
  row = rows[original_idx]
103
- del row["split"] # Remove original split column
104
- yield global_idx, row # Key is global shuffled index
 
39
  ]
40
  elif self.config.name == "filtered_random":
41
  filepath = dl_manager.download_and_extract("filtered.csv")
 
42
  with open(filepath, encoding="utf-8") as f:
43
  reader = csv.DictReader(f)
44
  rows = list(reader)
45
  n_total = len(rows)
 
46
  indices = list(range(n_total))
47
+ rng = random.Random(42)
48
  rng.shuffle(indices)
 
49
  n_train = int(0.8 * n_total)
50
  n_val = int(0.1 * n_total)
51
+ n_test = n_total - n_train - n_val
 
52
  return [
53
  datasets.SplitGenerator(
54
+ name=datasets.Split.TRAIN,
55
  gen_kwargs={
56
  "filepath": filepath,
57
  "shuffled_indices": indices,
 
60
  },
61
  ),
62
  datasets.SplitGenerator(
63
+ name=datasets.Split.VALIDATION,
64
  gen_kwargs={
65
  "filepath": filepath,
66
  "shuffled_indices": indices,
 
69
  },
70
  ),
71
  datasets.SplitGenerator(
72
+ name=datasets.Split.TEST,
73
  gen_kwargs={
74
  "filepath": filepath,
75
  "shuffled_indices": indices,
 
92
  del row["split"]
93
  yield idx, row
94
  elif self.config.name == "filtered_random":
 
95
  for global_idx in range(split_start, split_end):
96
  original_idx = shuffled_indices[global_idx]
97
  row = rows[original_idx]
98
+ del row["split"]
99
+ yield global_idx, row