zhuwq0 commited on
Commit
c8b078c
1 Parent(s): 37f6fec

rename data -> waveform

Browse files
Files changed (2) hide show
  1. example.py +22 -9
  2. quakeflow_nc.py +15 -19
example.py CHANGED
@@ -1,20 +1,33 @@
1
  # %%
 
2
  import numpy as np
3
- from datasets import load_dataset
4
  from torch.utils.data import DataLoader
5
 
6
- # quakeflow_nc = load_dataset("AI4EPS/quakeflow_nc", name="station_test", split="test")
7
- quakeflow_nc = load_dataset(
8
- "./quakeflow_nc.py",
9
- name="station_test",
10
- # name="event_test",
11
- split="test",
12
  download_mode="force_redownload",
 
 
13
  )
 
 
 
 
 
 
 
 
 
 
14
 
15
  # print the first sample of the iterable dataset
16
  for example in quakeflow_nc:
17
- print("\nIterable test\n")
 
18
  print(example.keys())
19
  for key in example.keys():
20
  if key == "data":
@@ -28,7 +41,7 @@ quakeflow_nc = quakeflow_nc.with_format("torch")
28
  dataloader = DataLoader(quakeflow_nc, batch_size=8, num_workers=0, collate_fn=lambda x: x)
29
 
30
  for batch in dataloader:
31
- print("\nDataloader test\n")
32
  print(f"Batch size: {len(batch)}")
33
  print(batch[0].keys())
34
  for key in batch[0].keys():
 
1
  # %%
2
+ import datasets
3
  import numpy as np
 
4
  from torch.utils.data import DataLoader
5
 
6
+ quakeflow_nc = datasets.load_dataset(
7
+ "AI4EPS/quakeflow_nc",
8
+ name="station",
9
+ split="train",
10
+ # name="station_test",
11
+ # split="test",
12
  download_mode="force_redownload",
13
+ trust_remote_code=True,
14
+ num_proc=36,
15
  )
16
+ # quakeflow_nc = datasets.load_dataset(
17
+ # "./quakeflow_nc.py",
18
+ # name="station",
19
+ # split="train",
20
+ # # name="statoin_test",
21
+ # # split="test",
22
+ # num_proc=36,
23
+ # )
24
+
25
+ print(quakeflow_nc)
26
 
27
  # print the first sample of the iterable dataset
28
  for example in quakeflow_nc:
29
+ print("\nIterable dataset\n")
30
+ print(example)
31
  print(example.keys())
32
  for key in example.keys():
33
  if key == "data":
 
41
  dataloader = DataLoader(quakeflow_nc, batch_size=8, num_workers=0, collate_fn=lambda x: x)
42
 
43
  for batch in dataloader:
44
+ print("\nDataloader dataset\n")
45
  print(f"Batch size: {len(batch)}")
46
  print(batch[0].keys())
47
  for key in batch[0].keys():
quakeflow_nc.py CHANGED
@@ -167,7 +167,7 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
167
  ):
168
  features = datasets.Features(
169
  {
170
- "data": datasets.Array2D(shape=(3, self.nt), dtype="float32"),
171
  "phase_time": datasets.Sequence(datasets.Value("string")),
172
  "phase_index": datasets.Sequence(datasets.Value("int32")),
173
  "phase_type": datasets.Sequence(datasets.Value("string")),
@@ -183,7 +183,7 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
183
  elif (self.config.name == "event") or (self.config.name == "event_train") or (self.config.name == "event_test"):
184
  features = datasets.Features(
185
  {
186
- "data": datasets.Array3D(shape=(None, 3, self.nt), dtype="float32"),
187
  "phase_time": datasets.Sequence(datasets.Sequence(datasets.Value("string"))),
188
  "phase_index": datasets.Sequence(datasets.Sequence(datasets.Value("int32"))),
189
  "phase_type": datasets.Sequence(datasets.Sequence(datasets.Value("string"))),
@@ -224,19 +224,17 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
224
  # By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
225
  urls = _URLS[self.config.name]
226
  # files = dl_manager.download(urls)
227
- files = dl_manager.download_and_extract(urls)
228
- # files = ["waveform_h5/1989.h5", "waveform_h5/1990.h5"]
229
- print(files)
 
230
 
231
  if self.config.name == "station" or self.config.name == "event":
232
  return [
233
  datasets.SplitGenerator(
234
  name=datasets.Split.TRAIN,
235
  # These kwargs will be passed to _generate_examples
236
- gen_kwargs={
237
- "filepath": files[:-1],
238
- "split": "train",
239
- },
240
  ),
241
  datasets.SplitGenerator(
242
  name=datasets.Split.TEST,
@@ -247,10 +245,7 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
247
  return [
248
  datasets.SplitGenerator(
249
  name=datasets.Split.TRAIN,
250
- gen_kwargs={
251
- "filepath": files,
252
- "split": "train",
253
- },
254
  ),
255
  ]
256
  elif self.config.name == "station_test" or self.config.name == "event_test":
@@ -269,6 +264,7 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
269
  # The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
270
 
271
  for file in filepath:
 
272
  with fsspec.open(file, "rb") as fs:
273
  with h5py.File(fs, "r") as fp:
274
  event_ids = list(fp.keys())
@@ -292,10 +288,10 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
292
  or (self.config.name == "station_train")
293
  or (self.config.name == "station_test")
294
  ):
295
- waveforms = np.zeros([3, self.nt], dtype="float32")
296
 
297
  for i, sta_id in enumerate(station_ids):
298
- waveforms[:, : self.nt] = event[sta_id][:, : self.nt]
299
  attrs = event[sta_id].attrs
300
  phase_type = attrs["phase_type"]
301
  phase_time = attrs["phase_time"]
@@ -304,7 +300,7 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
304
  station_location = [attrs["longitude"], attrs["latitude"], -attrs["elevation_m"] / 1e3]
305
 
306
  yield f"{event_id}/{sta_id}", {
307
- "data": waveforms,
308
  "phase_time": phase_time,
309
  "phase_index": phase_index,
310
  "phase_type": phase_type,
@@ -323,7 +319,7 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
323
  or (self.config.name == "event_test")
324
  ):
325
 
326
- waveforms = np.zeros([len(station_ids), 3, self.nt], dtype="float32")
327
  phase_type = []
328
  phase_time = []
329
  phase_index = []
@@ -331,7 +327,7 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
331
  station_location = []
332
 
333
  for i, sta_id in enumerate(station_ids):
334
- waveforms[i, :, : self.nt] = event[sta_id][:, : self.nt]
335
  attrs = event[sta_id].attrs
336
  phase_type.append(list(attrs["phase_type"]))
337
  phase_time.append(list(attrs["phase_time"]))
@@ -341,7 +337,7 @@ class QuakeFlow_NC(datasets.GeneratorBasedBuilder):
341
  [attrs["longitude"], attrs["latitude"], -attrs["elevation_m"] / 1e3]
342
  )
343
  yield event_id, {
344
- "data": waveforms,
345
  "phase_time": phase_time,
346
  "phase_index": phase_index,
347
  "phase_type": phase_type,
 
167
  ):
168
  features = datasets.Features(
169
  {
170
+ "waveform": datasets.Array2D(shape=(3, self.nt), dtype="float32"),
171
  "phase_time": datasets.Sequence(datasets.Value("string")),
172
  "phase_index": datasets.Sequence(datasets.Value("int32")),
173
  "phase_type": datasets.Sequence(datasets.Value("string")),
 
183
  elif (self.config.name == "event") or (self.config.name == "event_train") or (self.config.name == "event_test"):
184
  features = datasets.Features(
185
  {
186
+ "waveform": datasets.Array3D(shape=(None, 3, self.nt), dtype="float32"),
187
  "phase_time": datasets.Sequence(datasets.Sequence(datasets.Value("string"))),
188
  "phase_index": datasets.Sequence(datasets.Sequence(datasets.Value("int32"))),
189
  "phase_type": datasets.Sequence(datasets.Sequence(datasets.Value("string"))),
 
224
  # By default the archives will be extracted and a path to a cached folder where they are extracted is returned instead of the archive
225
  urls = _URLS[self.config.name]
226
  # files = dl_manager.download(urls)
227
+ # files = dl_manager.download_and_extract(urls)
228
+ files = [f"waveform_h5/{x}" for x in _FILES]
229
+ for file in sorted(files):
230
+ print(file)
231
 
232
  if self.config.name == "station" or self.config.name == "event":
233
  return [
234
  datasets.SplitGenerator(
235
  name=datasets.Split.TRAIN,
236
  # These kwargs will be passed to _generate_examples
237
+ gen_kwargs={"filepath": files[:-1], "split": "train"},
 
 
 
238
  ),
239
  datasets.SplitGenerator(
240
  name=datasets.Split.TEST,
 
245
  return [
246
  datasets.SplitGenerator(
247
  name=datasets.Split.TRAIN,
248
+ gen_kwargs={"filepath": files, "split": "train"},
 
 
 
249
  ),
250
  ]
251
  elif self.config.name == "station_test" or self.config.name == "event_test":
 
264
  # The `key` is for legacy reasons (tfds) and is not important in itself, but must be unique for each example.
265
 
266
  for file in filepath:
267
+ print(f"\nReading {file}")
268
  with fsspec.open(file, "rb") as fs:
269
  with h5py.File(fs, "r") as fp:
270
  event_ids = list(fp.keys())
 
288
  or (self.config.name == "station_train")
289
  or (self.config.name == "station_test")
290
  ):
291
+ waveform = np.zeros([3, self.nt], dtype="float32")
292
 
293
  for i, sta_id in enumerate(station_ids):
294
+ waveform[:, : self.nt] = event[sta_id][:, : self.nt]
295
  attrs = event[sta_id].attrs
296
  phase_type = attrs["phase_type"]
297
  phase_time = attrs["phase_time"]
 
300
  station_location = [attrs["longitude"], attrs["latitude"], -attrs["elevation_m"] / 1e3]
301
 
302
  yield f"{event_id}/{sta_id}", {
303
+ "waveform": waveform,
304
  "phase_time": phase_time,
305
  "phase_index": phase_index,
306
  "phase_type": phase_type,
 
319
  or (self.config.name == "event_test")
320
  ):
321
 
322
+ waveform = np.zeros([len(station_ids), 3, self.nt], dtype="float32")
323
  phase_type = []
324
  phase_time = []
325
  phase_index = []
 
327
  station_location = []
328
 
329
  for i, sta_id in enumerate(station_ids):
330
+ waveform[i, :, : self.nt] = event[sta_id][:, : self.nt]
331
  attrs = event[sta_id].attrs
332
  phase_type.append(list(attrs["phase_type"]))
333
  phase_time.append(list(attrs["phase_time"]))
 
337
  [attrs["longitude"], attrs["latitude"], -attrs["elevation_m"] / 1e3]
338
  )
339
  yield event_id, {
340
+ "waveform": waveform,
341
  "phase_time": phase_time,
342
  "phase_index": phase_index,
343
  "phase_type": phase_type,