def _create_datasets(self):
self.value = {}
max_sessions = 10 if self.fast else None
sessions = Session.fetch_all(only_real=True,
include_timeframes=True,
max=max_sessions)
pbar = tqdm(
total=self.count,
desc="Creating Datasets{}".format(" (fast)" if self.fast else ""))
for wl in self.window_lengths:
self.value[str(wl)] = {}
wl_datasets = list(
Session.full_dataset_gen(window_length=wl,
count=self.cv_splits,
sessions=sessions))
for sample_trim in self.sample_trims:
self.value[str(wl)][str(sample_trim)] = {}
st_datasets = [
ds.trim_none_seconds(sample_trim, return_copy=True)
for ds in wl_datasets
]
for ds_type in self.dataset_types:
dt_datasets = []
for st_ds in st_datasets:
ds = st_ds.copy()
ds = ds.reduced_dataset(ds_type)
ds = ds.normalize()
ds.shuffle()
dt_datasets.append(ds)
pbar.update(1)
self.value[str(wl)][str(sample_trim)][str(
ds_type)] = dt_datasets
pbar.close()
self.sample_shape = (2 * len(sample), cAshape[0])
if self.dim == 2:
coeffs = pywt.dwt2(sample, self.wavelet)
cA, (cH, cV, cD) = coeffs
res = np.vstack((cA, cH, cV, cD))
self.sample_shape = np.shape(res)
return self
if __name__ == '__main__':
ds = list(Session.full_dataset_gen(window_length=10))[0]
indices = []
for class_idx in [0, 1]:
for i, y in enumerate(ds.y):
if y == class_idx:
indices.append(i)
break
print(indices)
wavelet = DWT(dim=1, wavelet='db1')
wavelet.fit(ds.X)
for i in indices:
