Datasets:

ArXiv:
License:
gorold commited on
Commit
59e3fce
1 Parent(s): c93e81e

formatting

Browse files
Files changed (1) hide show
  1. cloudops_tsf.py +23 -12
cloudops_tsf.py CHANGED
@@ -269,19 +269,22 @@ class CloudOpsTSF(datasets.ArrowBasedBuilder):
269
  def _split_generators(self, dl_manager) -> list[datasets.SplitGenerator]:
270
  generators = []
271
  if self.config.train_test:
272
- downloaded_files = dl_manager.download_and_extract(f"{self.config.name}/train_test.zip")
 
 
273
  generators.append(
274
  datasets.SplitGenerator(
275
  name=TRAIN_TEST if self.config.train_test else PRETRAIN,
276
- gen_kwargs={"filepath": downloaded_files}
277
  )
278
  )
279
  if self.config.pretrain:
280
- downloaded_files = dl_manager.download_and_extract(f"{self.config.name}/pretrain.zip")
 
 
281
  generators.append(
282
  datasets.SplitGenerator(
283
- name=PRETRAIN,
284
- gen_kwargs={"filepath": downloaded_files}
285
  )
286
  )
287
  return generators
@@ -294,13 +297,21 @@ class CloudOpsTSF(datasets.ArrowBasedBuilder):
294
  schema = batch.schema
295
  if self.config.include_metadata:
296
  freq = pa.array([self.config.freq] * len(batch))
297
- prediction_length = pa.array([self.config.prediction_length] * len(batch))
298
- rolling_evaluations = pa.array([self.config.rolling_evaluations] * len(batch))
 
 
 
 
299
  stride = pa.array([self.config.stride] * len(batch))
300
  columns += [freq, prediction_length, rolling_evaluations, stride]
301
- for pa_field in [pa.field('freq', pa.string()),
302
- pa.field('prediction_length', pa.int32()),
303
- pa.field('rolling_evaluations', pa.int32()),
304
- pa.field('stride', pa.int32())]:
 
 
305
  schema = schema.append(pa_field)
306
- yield batch[FieldName.ITEM_ID].to_pylist(), pa.Table.from_arrays(columns, schema=schema)
 
 
 
269
  def _split_generators(self, dl_manager) -> list[datasets.SplitGenerator]:
270
  generators = []
271
  if self.config.train_test:
272
+ downloaded_files = dl_manager.download_and_extract(
273
+ f"{self.config.name}/train_test.zip"
274
+ )
275
  generators.append(
276
  datasets.SplitGenerator(
277
  name=TRAIN_TEST if self.config.train_test else PRETRAIN,
278
+ gen_kwargs={"filepath": downloaded_files},
279
  )
280
  )
281
  if self.config.pretrain:
282
+ downloaded_files = dl_manager.download_and_extract(
283
+ f"{self.config.name}/pretrain.zip"
284
+ )
285
  generators.append(
286
  datasets.SplitGenerator(
287
+ name=PRETRAIN, gen_kwargs={"filepath": downloaded_files}
 
288
  )
289
  )
290
  return generators
 
297
  schema = batch.schema
298
  if self.config.include_metadata:
299
  freq = pa.array([self.config.freq] * len(batch))
300
+ prediction_length = pa.array(
301
+ [self.config.prediction_length] * len(batch)
302
+ )
303
+ rolling_evaluations = pa.array(
304
+ [self.config.rolling_evaluations] * len(batch)
305
+ )
306
  stride = pa.array([self.config.stride] * len(batch))
307
  columns += [freq, prediction_length, rolling_evaluations, stride]
308
+ for pa_field in [
309
+ pa.field("freq", pa.string()),
310
+ pa.field("prediction_length", pa.int32()),
311
+ pa.field("rolling_evaluations", pa.int32()),
312
+ pa.field("stride", pa.int32()),
313
+ ]:
314
  schema = schema.append(pa_field)
315
+ yield batch[FieldName.ITEM_ID].to_pylist(), pa.Table.from_arrays(
316
+ columns, schema=schema
317
+ )