def test_timeout_1(hdf5, lt_ctx):
with mock.patch('h5py.File.visititems', side_effect=TimeoutError("too slow")):
params = H5DataSet.detect_params(hdf5.filename, executor=lt_ctx.executor)["parameters"]
assert list(params.keys()) == ["path"]
ds = H5DataSet(
path=hdf5.filename, ds_path="data",
)
ds = ds.initialize(lt_ctx.executor)
diags = ds.diagnostics
print(diags)
def test_timeout_2(hdf5):
with mock.patch('time.time', side_effect=[1, 30, 30, 60]):
params = H5DataSet.detect_params(hdf5.filename)
assert list(params.keys()) == ["path"]
ds = H5DataSet(path=hdf5.filename,
ds_path="data",
tileshape=(1, 4, 16, 16))
ds = ds.initialize()
diags = ds.diagnostics
print(diags)
def test_timeout_1(hdf5):
with mock.patch('h5py.File.visititems',
side_effect=TimeoutError("too slow")):
params = H5DataSet.detect_params(hdf5.filename)
assert list(params.keys()) == ["path"]
ds = H5DataSet(path=hdf5.filename,
ds_path="data",
tileshape=(1, 4, 16, 16))
ds = ds.initialize()
diags = ds.diagnostics
print(diags)
def test_timeout_2(hdf5, lt_ctx):
print(threading.enumerate())
with mock.patch('libertem.io.dataset.hdf5.current_time', side_effect=[1, 30]):
params = H5DataSet.detect_params(hdf5.filename, executor=lt_ctx.executor)["parameters"]
assert list(params.keys()) == ["path"]
ds = H5DataSet(
path=hdf5.filename, ds_path="data",
)
ds = ds.initialize(lt_ctx.executor)
print(threading.enumerate())
with mock.patch('libertem.io.dataset.hdf5.current_time', side_effect=[30, 60]):
diags = ds.diagnostics
print(diags)
def do_com(fn, tileshape):
ds = H5DataSet(
path=fn,
ds_path="data",
tileshape=tileshape,
target_size=512*1024*1024,
)
masks = [
# summation of all pixels:
lambda: np.ones(shape=ds.shape[2:]),
# gradient from left to right
lambda: gradient_x(*ds.shape[2:]),
# gradient from top to bottom
lambda: gradient_y(*ds.shape[2:]),
]
job = ApplyMasksJob(dataset=ds, mask_factories=masks)
print(job.masks.computed_masks)
print("\n\n")
executor = InlineJobExecutor()
full_result = np.zeros(shape=(3,) + ds.shape[:2])
color = np.zeros(shape=(3,) + ds.shape[:2])
for result in executor.run_job(job):
for tile in result:
print(tile)
print(tile.data[0])
color[tile.tile_slice.get()[:2]] += 1
tile.copy_to_result(full_result)
x_centers = np.divide(full_result[1], full_result[0])
y_centers = np.divide(full_result[2], full_result[0])
print(color)
return full_result, x_centers, y_centers
def test_diags(hdf5, lt_ctx):
ds = H5DataSet(
path=hdf5.filename,
ds_path="data",
)
ds = ds.initialize(lt_ctx.executor)
print(ds.diagnostics)
def hdf5_ds_1(hdf5):
ds = H5DataSet(
path=hdf5.filename,
ds_path="data",
)
ds = ds.initialize(InlineJobExecutor())
return ds
def test_check_valid(hdf5, lt_ctx):
ds = H5DataSet(
path=hdf5.filename,
ds_path="data",
)
ds = ds.initialize(lt_ctx.executor)
assert ds.check_valid()
def hdf5_ds_1(hdf5):
ds = H5DataSet(path=hdf5.filename,
ds_path="data",
tileshape=(1, 5, 16, 16),
target_size=512 * 1024 * 1024)
ds.initialize()
return ds
def hdf5_ds_large_sig(random_hdf5):
ds = H5DataSet(
path=random_hdf5.filename,
ds_path="data",
)
ds = ds.initialize(InlineJobExecutor())
return ds
def test_roi_3(hdf5, lt_ctx):
ds = H5DataSet(
path=hdf5.filename, ds_path="data",
target_size=12800*2,
)
ds = ds.initialize(lt_ctx.executor)
roi = np.zeros(ds.shape.flatten_nav().nav, dtype=bool)
roi[24] = 1
tileshape = Shape(
(16,) + tuple(ds.shape.sig),
sig_dims=ds.shape.sig.dims
)
tiling_scheme = TilingScheme.make_for_shape(
tileshape=tileshape,
dataset_shape=ds.shape,
)
tiles = []
for p in ds.get_partitions():
for tile in p.get_tiles(tiling_scheme=tiling_scheme, dest_dtype="float32", roi=roi):
print("tile:", tile)
tiles.append(tile)
assert len(tiles) == 1
assert tiles[0].tile_slice.shape.nav.size == 1
assert tuple(tiles[0].tile_slice.shape.sig) == (16, 16)
assert tiles[0].tile_slice.origin == (0, 0, 0)
assert np.allclose(tiles[0].data, hdf5['data'][4, 4])
def test_cloudpickle(lt_ctx, hdf5):
ds = H5DataSet(
path=hdf5.filename, ds_path="data", target_size=512*1024*1024
)
pickled = cloudpickle.dumps(ds)
loaded = cloudpickle.loads(pickled)
assert loaded._dtype is None
assert loaded._shape is None
repr(loaded)
ds = ds.initialize(lt_ctx.executor)
pickled = cloudpickle.dumps(ds)
loaded = cloudpickle.loads(pickled)
assert loaded._dtype is not None
assert loaded._shape is not None
loaded.shape
loaded.dtype
repr(loaded)
# let's keep the pickled dataset size small-ish:
assert len(pickled) < 1 * 1024
def test_cache_key_json_serializable(hdf5, lt_ctx):
ds = H5DataSet(
path=hdf5.filename,
ds_path="data",
)
ds = ds.initialize(lt_ctx.executor)
json.dumps(ds.get_cache_key())
def test_read_2(lt_ctx, hdf5):
ds = H5DataSet(path=hdf5.filename,
ds_path="data",
tileshape=(1, 3, 16, 16))
ds = ds.initialize(lt_ctx.executor)
for p in ds.get_partitions():
for t in p.get_tiles():
print(t.tile_slice)
def create_random_hdf5(path):
with h5py.File(path, 'w') as f:
sample_data = np.random.randn(16, 16, 16, 16).astype("float32")
f.create_dataset("data", (16, 16, 16, 16), data=sample_data)
# read and provide the ds
ds = H5DataSet(path=path, ds_path='data')
ds = ds.initialize(InlineJobExecutor())
return ds
def test_pick(hdf5, lt_ctx):
ds = H5DataSet(path=hdf5.filename,
ds_path="data",
tileshape=(1, 3, 16, 16))
ds.initialize()
assert len(ds.shape) == 4
print(ds.shape)
pick = lt_ctx.create_pick_analysis(dataset=ds, x=2, y=3)
lt_ctx.run(pick)
def test_pick(hdf5, lt_ctx, TYPE):
ds = H5DataSet(
path=hdf5.filename, ds_path="data",
)
ds = ds.initialize(lt_ctx.executor)
assert len(ds.shape) == 4
print(ds.shape)
pick = lt_ctx.create_pick_analysis(dataset=ds, x=2, y=3)
pick.TYPE = TYPE
lt_ctx.run(pick)
def test_read_3(lt_ctx, random_hdf5):
# try with smaller partitions:
ds = H5DataSet(path=random_hdf5.filename,
ds_path="data",
tileshape=(1, 2, 16, 16),
target_size=4096)
ds = ds.initialize(lt_ctx.executor)
for p in ds.get_partitions():
for t in p.get_tiles():
print(t.tile_slice)
def test_auto_tileshape(chunked_hdf5, lt_ctx):
ds = H5DataSet(
path=chunked_hdf5.filename,
ds_path="data",
)
ds = ds.initialize(lt_ctx.executor)
p = next(ds.get_partitions())
t = next(p.get_tiles(dest_dtype="float32", target_size=4 * 1024))
assert tuple(p._get_tileshape("float32", 4 * 1024)) == (1, 4, 16, 16)
assert tuple(t.tile_slice.shape) == (4, 16, 16)
def test_read_3(lt_ctx, random_hdf5):
# try with smaller partitions:
ds = H5DataSet(path=random_hdf5.filename, ds_path="data", target_size=4096)
ds = ds.initialize(lt_ctx.executor)
tileshape = Shape((16, ) + tuple(ds.shape.sig), sig_dims=ds.shape.sig.dims)
tiling_scheme = TilingScheme.make_for_shape(
tileshape=tileshape,
dataset_shape=ds.shape,
)
for p in ds.get_partitions():
for t in p.get_tiles(tiling_scheme=tiling_scheme):
print(t.tile_slice)
def test_read_2(lt_ctx, hdf5):
ds = H5DataSet(
path=hdf5.filename,
ds_path="data",
)
ds = ds.initialize(lt_ctx.executor)
tileshape = Shape((16, ) + tuple(ds.shape.sig), sig_dims=ds.shape.sig.dims)
tiling_scheme = TilingScheme.make_for_shape(
tileshape=tileshape,
dataset_shape=ds.shape,
)
for p in ds.get_partitions():
for t in p.get_tiles(tiling_scheme=tiling_scheme):
print(t.tile_slice)
def test_roi_4(hdf5, lt_ctx):
ds = H5DataSet(
path=hdf5.filename,
ds_path="data",
target_size=12800 * 2,
)
ds = ds.initialize(lt_ctx.executor)
roi = np.random.choice(size=ds.shape.flatten_nav().nav, a=[True, False])
sum_udf = lt_ctx.create_sum_analysis(dataset=ds)
sumres = lt_ctx.run(sum_udf, roi=roi)['intensity']
assert np.allclose(
sumres, np.sum(hdf5['data'][:].reshape(25, 16, 16)[roi, ...], axis=0))
def test_roi_2(random_hdf5, lt_ctx, mnp):
ds = H5DataSet(
path=random_hdf5.filename,
ds_path="data",
tileshape=(1, 4, 16, 16),
min_num_partitions=mnp,
)
ds = ds.initialize(lt_ctx.executor)
roi = {
"shape": "disk",
"cx": 2,
"cy": 2,
"r": 1,
}
analysis = SumAnalysis(dataset=ds, parameters={
"roi": roi,
})
print(analysis.get_roi())
results = lt_ctx.run(analysis)
# let's draw a circle!
mask = np.full((5, 5), False)
mask[1, 2] = True
mask[2, 1:4] = True
mask[3, 2] = True
print(mask)
assert mask.shape == (5, 5)
assert mask.dtype == np.bool
reader = ds.get_reader()
with reader.get_h5ds() as h5ds:
data = np.array(h5ds)
# applying the mask flattens the first two dimensions, so we
# only sum over axis 0 here:
expected = data[mask, ...].sum(axis=(0, ))
assert expected.shape == (16, 16)
assert results.intensity.raw_data.shape == (16, 16)
# is not equal to results without mask:
assert not np.allclose(results.intensity.raw_data,
data.sum(axis=(0, 1)))
# ... but rather like `expected`:
assert np.allclose(results.intensity.raw_data, expected)
def test_roi_5(hdf5, lt_ctx):
ds = H5DataSet(
path=hdf5.filename, ds_path="data",
target_size=12800*2,
)
ds = ds.initialize(lt_ctx.executor)
roi = np.random.choice(size=ds.shape.flatten_nav().nav, a=[True, False])
udf = SumSigUDF()
sumres = lt_ctx.run_udf(dataset=ds, udf=udf, roi=roi)['intensity']
assert np.allclose(
sumres.raw_data,
np.sum(hdf5['data'][:][roi.reshape(5, 5), ...], axis=(1, 2))
)
def test_roi_1(hdf5, lt_ctx):
ds = H5DataSet(path=hdf5.filename,
ds_path="data",
tileshape=(1, 4, 16, 16))
ds = ds.initialize(lt_ctx.executor)
p = next(ds.get_partitions())
roi = np.zeros(p.meta.shape.flatten_nav().nav, dtype=bool)
roi[0] = 1
tiles = []
for tile in p.get_tiles(dest_dtype="float32", roi=roi):
print("tile:", tile)
tiles.append(tile)
assert len(tiles) == 1
assert tiles[0].tile_slice.shape.nav.size == 1
assert tuple(tiles[0].tile_slice.shape.sig) == (16, 16)
assert tiles[0].tile_slice.origin == (0, 0, 0)
def test_scheme_idx(lt_ctx, hdf5):
ds = H5DataSet(
path=hdf5.filename,
ds_path="data",
)
ds = ds.initialize(lt_ctx.executor)
p = next(ds.get_partitions())
sig_shape = tuple(ds.shape.sig)
tileshape = Shape((16, ) + sig_shape[:-1] + (sig_shape[-1] // 2, ),
sig_dims=ds.shape.sig.dims)
tiling_scheme = TilingScheme.make_for_shape(
tileshape=tileshape,
dataset_shape=ds.shape,
)
tiles = p.get_tiles(tiling_scheme=tiling_scheme)
for tile, expected_idx in zip(tiles, itertools.cycle([0, 1])):
print(tile.scheme_idx, tile.tile_slice)
assert tile.scheme_idx == expected_idx
def test_roi_1(hdf5, lt_ctx):
ds = H5DataSet(
path=hdf5.filename, ds_path="data",
)
ds = ds.initialize(lt_ctx.executor)
p = next(ds.get_partitions())
roi = np.zeros(p.meta.shape.flatten_nav().nav, dtype=bool)
roi[0] = 1
tiles = []
tileshape = Shape(
(16,) + tuple(ds.shape.sig),
sig_dims=ds.shape.sig.dims
)
tiling_scheme = TilingScheme.make_for_shape(
tileshape=tileshape,
dataset_shape=ds.shape,
)
for tile in p.get_tiles(tiling_scheme=tiling_scheme, dest_dtype="float32", roi=roi):
print("tile:", tile)
tiles.append(tile)
assert len(tiles) == 1
assert tiles[0].tile_slice.shape.nav.size == 1
assert tuple(tiles[0].tile_slice.shape.sig) == (16, 16)
assert tiles[0].tile_slice.origin == (0, 0, 0)
def hdf5_ds_1(hdf5):
ds = H5DataSet(path=hdf5.filename,
ds_path="data",
tileshape=(1, 5, 16, 16))
ds = ds.initialize()
return ds
def hdf5_ds_2(random_hdf5):
ds = H5DataSet(
path=random_hdf5.filename, ds_path="data", tileshape=(1, 5, 16, 16)
)
ds = ds.initialize(InlineJobExecutor())
return ds
def hdf5_ds_5d(hdf5_5d):
ds = H5DataSet(path=hdf5_5d.filename,
ds_path="data",
tileshape=(1, 1, 1, 16, 16))
ds = ds.initialize(InlineJobExecutor())
return ds
