def test_write_to_any_location(temp_h5_file):
loc = "path/path_1"
class ContainerA(Container):
a = Field(0, "some integer field")
a = ContainerA()
a.a = 1
with HDF5TableWriter(temp_h5_file, group_name=loc + "/group_1") as h5:
for _ in range(5):
h5.write("table", a)
h5.write("deeper/table2", a)
with HDF5TableReader(temp_h5_file) as h5:
for a in h5.read("/" + loc + "/group_1/table", ContainerA()):
assert a.a == 1
with HDF5TableReader(temp_h5_file) as h5:
for a in h5.read("/" + loc + "/group_1/deeper/table2", ContainerA()):
assert a.a == 1
def test_read_multiple_containers():
hillas_parameter_container = HillasParametersContainer(x=1 * u.m,
y=1 * u.m,
length=1 * u.m,
width=1 * u.m)
leakage_container = LeakageContainer(
pixels_width_1=0.1,
pixels_width_2=0.1,
intensity_width_1=0.1,
intensity_width_2=0.1,
)
with tempfile.NamedTemporaryFile() as f:
with HDF5TableWriter(f.name, group_name="dl1",
add_prefix=True) as writer:
writer.write("params",
[hillas_parameter_container, leakage_container])
df = pd.read_hdf(f.name, key="/dl1/params")
assert "hillas_x" in df.columns
assert "leakage_pixels_width_1" in df.columns
# test reading both containers separately
with HDF5TableReader(f.name) as reader:
generator = reader.read("/dl1/params",
HillasParametersContainer(),
prefixes=True)
hillas = next(generator)
for value, read_value in zip(
hillas_parameter_container.as_dict().values(),
hillas.as_dict().values()):
np.testing.assert_equal(value, read_value)
with HDF5TableReader(f.name) as reader:
generator = reader.read("/dl1/params",
LeakageContainer(),
prefixes=True)
leakage = next(generator)
for value, read_value in zip(leakage_container.as_dict().values(),
leakage.as_dict().values()):
np.testing.assert_equal(value, read_value)
# test reading both containers simultaneously
with HDF5TableReader(f.name) as reader:
generator = reader.read(
"/dl1/params",
[HillasParametersContainer(),
LeakageContainer()],
prefixes=True,
)
hillas_, leakage_ = next(generator)
for value, read_value in zip(leakage_container.as_dict().values(),
leakage_.as_dict().values()):
np.testing.assert_equal(value, read_value)
for value, read_value in zip(
hillas_parameter_container.as_dict().values(),
hillas_.as_dict().values()):
np.testing.assert_equal(value, read_value)
def test_read_without_prefixes(tmp_path):
path = tmp_path / "test.h5"
hillas_parameter_container = HillasParametersContainer(fov_lon=1 * u.deg,
fov_lat=1 * u.deg,
length=1 * u.deg,
width=1 * u.deg)
leakage_container = LeakageContainer(
pixels_width_1=0.1,
pixels_width_2=0.1,
intensity_width_1=0.1,
intensity_width_2=0.1,
)
with HDF5TableWriter(path, group_name="dl1", add_prefix=False) as writer:
writer.write("params", [hillas_parameter_container, leakage_container])
df = pd.read_hdf(path, key="/dl1/params")
assert "fov_lon" in df.columns
assert "pixels_width_1" in df.columns
# call with prefixes=False
with HDF5TableReader(path) as reader:
generator = reader.read(
"/dl1/params",
[HillasParametersContainer(),
LeakageContainer()],
prefixes=False,
)
hillas_, leakage_ = next(generator)
for value, read_value in zip(leakage_container.as_dict().values(),
leakage_.as_dict().values()):
np.testing.assert_equal(value, read_value)
for value, read_value in zip(hillas_parameter_container.as_dict().values(),
hillas_.as_dict().values()):
np.testing.assert_equal(value, read_value)
# call with manually removed prefixes
with HDF5TableReader(path) as reader:
generator = reader.read(
"/dl1/params",
[
HillasParametersContainer(prefix=""),
LeakageContainer(prefix="")
],
prefixes=True,
)
hillas_, leakage_ = next(generator)
for value, read_value in zip(leakage_container.as_dict().values(),
leakage_.as_dict().values()):
np.testing.assert_equal(value, read_value)
for value, read_value in zip(hillas_parameter_container.as_dict().values(),
hillas_.as_dict().values()):
np.testing.assert_equal(value, read_value)
def test_read_multiple_containers(tmp_path):
path = tmp_path / "test_append.h5"
hillas_parameter_container = HillasParametersContainer(fov_lon=1 * u.deg,
fov_lat=1 * u.deg,
length=1 * u.deg,
width=1 * u.deg)
leakage_container = LeakageContainer(
pixels_width_1=0.1,
pixels_width_2=0.1,
intensity_width_1=0.1,
intensity_width_2=0.1,
)
with HDF5TableWriter(path, group_name="dl1", add_prefix=True) as writer:
writer.write("params", [hillas_parameter_container, leakage_container])
df = pd.read_hdf(path, key="/dl1/params")
assert "hillas_fov_lon" in df.columns
assert "leakage_pixels_width_1" in df.columns
# test reading both containers separately
with HDF5TableReader(path) as reader:
generator = reader.read("/dl1/params",
HillasParametersContainer(),
prefixes=True)
hillas = next(generator)
for value, read_value in zip(hillas_parameter_container.as_dict().values(),
hillas.as_dict().values()):
np.testing.assert_equal(value, read_value)
with HDF5TableReader(path) as reader:
generator = reader.read("/dl1/params",
LeakageContainer(),
prefixes=True)
leakage = next(generator)
for value, read_value in zip(leakage_container.as_dict().values(),
leakage.as_dict().values()):
np.testing.assert_equal(value, read_value)
# test reading both containers simultaneously
with HDF5TableReader(path) as reader:
generator = reader.read(
"/dl1/params",
[HillasParametersContainer(),
LeakageContainer()],
prefixes=True,
)
hillas_, leakage_ = next(generator)
for value, read_value in zip(leakage_container.as_dict().values(),
leakage_.as_dict().values()):
np.testing.assert_equal(value, read_value)
for value, read_value in zip(hillas_parameter_container.as_dict().values(),
hillas_.as_dict().values()):
np.testing.assert_equal(value, read_value)
def test_read_without_prefixes():
hillas_parameter_container = HillasParametersContainer(
x=1 * u.m, y=1 * u.m, length=1 * u.m, width=1 * u.m
)
leakage_container = LeakageContainer(
pixels_width_1=0.1,
pixels_width_2=0.1,
intensity_width_1=0.1,
intensity_width_2=0.1,
)
with tempfile.NamedTemporaryFile() as f:
with HDF5TableWriter(f.name, group_name="dl1", add_prefix=False) as writer:
writer.write("params", [hillas_parameter_container, leakage_container])
df = pd.read_hdf(f.name, key="/dl1/params")
assert "x" in df.columns
assert "pixels_width_1" in df.columns
# call with prefixes=False
with HDF5TableReader(f.name) as reader:
generator = reader.read(
"/dl1/params",
[HillasParametersContainer(), LeakageContainer()],
prefixes=False,
)
hillas_, leakage_ = next(generator)
for value, read_value in zip(
leakage_container.as_dict().values(), leakage_.as_dict().values()
):
np.testing.assert_equal(value, read_value)
for value, read_value in zip(
hillas_parameter_container.as_dict().values(), hillas_.as_dict().values()
):
np.testing.assert_equal(value, read_value)
# call with manually removed prefixes
with HDF5TableReader(f.name) as reader:
generator = reader.read(
"/dl1/params",
[HillasParametersContainer(prefix=""), LeakageContainer(prefix="")],
prefixes=True,
)
hillas_, leakage_ = next(generator)
for value, read_value in zip(
leakage_container.as_dict().values(), leakage_.as_dict().values()
):
np.testing.assert_equal(value, read_value)
for value, read_value in zip(
hillas_parameter_container.as_dict().values(), hillas_.as_dict().values()
):
np.testing.assert_equal(value, read_value)
def test_column_transforms(tmp_path):
""" ensure a user-added column transform is applied """
tmp_file = tmp_path / "test_column_transforms.hdf5"
class SomeContainer(Container):
value = Field(-1, "some value that should be transformed")
cont = SomeContainer()
def my_transform(x):
""" makes a length-3 array from x"""
return np.ones(3) * x
with HDF5TableWriter(tmp_file, group_name="data") as writer:
# add user generated transform for the "value" column
cont.value = 6.0
writer.add_column_transform("mytable", "value", my_transform)
writer.write("mytable", cont)
# check that we get a length-3 array when reading back
with HDF5TableReader(tmp_file, mode="r") as reader:
for data in reader.read("/data/mytable", SomeContainer()):
print(data)
assert data.value.shape == (3, )
assert np.allclose(data.value, [6.0, 6.0, 6.0])
def test_read_container(temp_h5_file):
r0tel1 = R0CameraContainer()
r0tel2 = R0CameraContainer()
mc = MCEventContainer()
with HDF5TableReader(temp_h5_file) as reader:
# get the generators for each table
mctab = reader.read("/R0/MC", mc)
r0tab1 = reader.read("/R0/tel_001", r0tel1)
r0tab2 = reader.read("/R0/tel_002", r0tel2)
# read all 3 tables in sync
for ii in range(3):
m = next(mctab)
r0_1 = next(r0tab1)
r0_2 = next(r0tab2)
print("MC:", m)
print("t0:", r0_1.waveform)
print("t1:", r0_2.waveform)
print("---------------------------")
assert "test_attribute" in r0_1.meta
assert r0_1.meta["date"] == "2020-10-10"
def test_append_mode(temp_h5_file):
class ContainerA(Container):
a = Field(int)
a = ContainerA()
a.a = 1
# First open with 'w' mode to clear the file and add a Container
with HDF5TableWriter(temp_h5_file, 'group') as h5:
h5.write('table_1', a)
# Try to append A again
with HDF5TableWriter(temp_h5_file, 'group', mode='a') as h5:
h5.write('table_2', a)
# Check if file has two tables with a = 1
with HDF5TableReader(temp_h5_file) as h5:
for a in h5.read('/group/table_1', ContainerA()):
assert a.a == 1
for a in h5.read('/group/table_2', ContainerA()):
assert a.a == 1
def test_read_duplicated_container_types(tmp_path):
path = tmp_path / "test.h5"
hillas_config_1 = HillasParametersContainer(
x=1 * u.m, y=2 * u.m, length=3 * u.m, width=4 * u.m, prefix="hillas_1"
)
hillas_config_2 = HillasParametersContainer(
x=2 * u.m, y=3 * u.m, length=4 * u.m, width=5 * u.m, prefix="hillas_2"
)
with HDF5TableWriter(path, group_name="dl1", add_prefix=True) as writer:
writer.write("params", [hillas_config_1, hillas_config_2])
df = pd.read_hdf(path, key="/dl1/params")
assert "hillas_1_x" in df.columns
assert "hillas_2_x" in df.columns
with HDF5TableReader(path) as reader:
generator = reader.read(
"/dl1/params",
[HillasParametersContainer(), HillasParametersContainer()],
prefixes=["hillas_1", "hillas_2"],
)
hillas_1, hillas_2 = next(generator)
for value, read_value in zip(
hillas_config_1.as_dict().values(), hillas_1.as_dict().values()
):
np.testing.assert_equal(value, read_value)
for value, read_value in zip(
hillas_config_2.as_dict().values(), hillas_2.as_dict().values()
):
np.testing.assert_equal(value, read_value)
def test_fixed_point_column_transform(tmp_path):
""" ensure a user-added column transform is applied """
from ctapipe.io.tableio import FixedPointColumnTransform
tmp_file = tmp_path / "test_column_transforms.hdf5"
class SomeContainer(Container):
container_prefix = ""
image = Field(np.array([np.nan, np.inf, -np.inf]))
cont = SomeContainer()
with HDF5TableWriter(tmp_file, group_name="data") as writer:
writer.add_column_transform(
"signed", "image",
FixedPointColumnTransform(100, 0, np.float64, np.int32))
writer.add_column_transform(
"unsigned",
"image",
FixedPointColumnTransform(100, 0, np.float64, np.uint32),
)
# add user generated transform for the "value" column
writer.write("signed", cont)
writer.write("unsigned", cont)
with HDF5TableReader(tmp_file, mode="r") as reader:
signed = next(reader.read("/data/signed", SomeContainer()))
unsigned = next(reader.read("/data/unsigned", SomeContainer()))
for data in (signed, unsigned):
# check we get our original nans back
assert np.isnan(data.image[0])
assert np.isposinf(data.image[1])
assert np.isneginf(data.image[2])
def test_column_transforms_regexps(tmp_path):
""" ensure a user-added column transform is applied when given as a regexp"""
tmp_file = tmp_path / "test_column_transforms.hdf5"
def multiply_by_10(x):
return x * 10
class SomeContainer(Container):
container_prefix = ""
hillas_x = Field(1)
hillas_y = Field(1)
cont = SomeContainer()
with HDF5TableWriter(tmp_file, group_name="data") as writer:
writer.add_column_transform_regexp("my.*", "hillas_.*", multiply_by_10)
writer.add_column_transform_regexp("anothertable", "hillas_y",
multiply_by_10)
writer.write("mytable", cont)
writer.write("anothertable", cont)
# check that we get back the transformed values (note here a round trip will
# not work, as there is no inverse transform in this test)
with HDF5TableReader(tmp_file, mode="r") as reader:
data = next(reader.read("/data/mytable", SomeContainer()))
assert data.hillas_x == 10
assert data.hillas_y == 10
data = next(reader.read("/data/anothertable", SomeContainer()))
assert data.hillas_x == 1
assert data.hillas_y == 10
def test_column_transforms(tmp_path):
""" ensure a user-added column transform is applied """
from ctapipe.containers import NAN_TIME
from ctapipe.io.tableio import FixedPointColumnTransform
tmp_file = tmp_path / "test_column_transforms.hdf5"
class SomeContainer(Container):
container_prefix = ""
current = Field(1 * u.A, unit=u.uA)
time = Field(NAN_TIME)
image = Field(np.array([1.234, 123.456]))
cont = SomeContainer()
with HDF5TableWriter(tmp_file, group_name="data") as writer:
writer.add_column_transform(
"mytable", "image",
FixedPointColumnTransform(100, 0, np.float64, np.int32))
# add user generated transform for the "value" column
writer.write("mytable", cont)
# check that we get a length-3 array when reading back
with HDF5TableReader(tmp_file, mode="r") as reader:
data = next(reader.read("/data/mytable", SomeContainer()))
assert data.current.value == 1e6
assert data.current.unit == u.uA
assert isinstance(data.time, Time)
assert data.time == NAN_TIME
# rounded to two digits
assert np.all(data.image == np.array([1.23, 123.45]))
def test_read_container(temp_h5_file):
r0tel1 = R0CameraContainer()
r0tel2 = R0CameraContainer()
mc = MCEventContainer()
reader = HDF5TableReader(str(temp_h5_file))
# get the generators for each table
mctab = reader.read('/R0/MC', mc)
r0tab1 = reader.read('/R0/tel_001', r0tel1)
r0tab2 = reader.read('/R0/tel_002', r0tel2)
# read all 3 tables in sync
for ii in range(3):
m = next(mctab)
r0_1 = next(r0tab1)
r0_2 = next(r0tab2)
print("MC:", m)
print("t0:", r0_1.image)
print("t1:", r0_2.image)
print("---------------------------")
assert 'test_attribute' in r0_1.meta
assert r0_1.meta['date'] == "2020-10-10"
reader.close()
def test_reader_closes_file(temp_h5_file):
with HDF5TableReader(str(temp_h5_file)) as h5_table:
assert h5_table._h5file.isopen == True
assert h5_table._h5file.isopen == False
def test_reader_closes_file(temp_h5_file):
with HDF5TableReader(temp_h5_file) as h5_table:
assert h5_table._h5file.isopen == 1
assert h5_table._h5file.isopen == 0
def test_read_container(temp_h5_file):
r0tel1 = R0CameraContainer()
r0tel2 = R0CameraContainer()
sim_shower = SimulatedShowerContainer()
with HDF5TableReader(temp_h5_file) as reader:
# get the generators for each table
# test supplying a single container as well as an
# iterable with one entry only
simtab = reader.read("/R0/sim_shower", (sim_shower, ))
r0tab1 = reader.read("/R0/tel_001", r0tel1)
r0tab2 = reader.read("/R0/tel_002", r0tel2)
# read all 3 tables in sync
for ii in range(3):
m = next(simtab)[0]
r0_1 = next(r0tab1)
r0_2 = next(r0tab2)
print("sim_shower:", m)
print("t0:", r0_1.waveform)
print("t1:", r0_2.waveform)
print("---------------------------")
assert "test_attribute" in r0_1.meta
assert r0_1.meta["date"] == "2020-10-10"
def _initialize_correction(self):
"""
Read the correction from hdf5 calibration file
"""
self.mon_data.tels_with_data = self.allowed_tels
self.log.info(f"read {self.calibration_path}")
try:
with HDF5TableReader(self.calibration_path) as h5_table:
assert h5_table._h5file.isopen == True
for telid in self.allowed_tels:
# read the calibration data for the moment only one event
table = '/tel_' + str(telid) + '/calibration'
next(
h5_table.read(table,
self.mon_data.tel[telid].calibration))
# eliminate inf values (should be done probably before)
dc_to_pe = self.mon_data.tel[telid].calibration.dc_to_pe
dc_to_pe[np.isinf(dc_to_pe)] = 0
self.log.info(
f"read {self.mon_data.tel[telid].calibration.dc_to_pe}"
)
except:
self.log.error(
f"Problem in reading calibration file {self.calibration_path}")
def test_read_whole_table(temp_h5_file):
sim_shower = SimulatedShowerContainer()
with HDF5TableReader(temp_h5_file) as reader:
for cont in reader.read("/R0/sim_shower", sim_shower):
print(cont)
def test_read_whole_table(temp_h5_file):
mc = MCEventContainer()
with HDF5TableReader(temp_h5_file) as reader:
for cont in reader.read("/R0/MC", mc):
print(cont)
def test_read_whole_table(temp_h5_file):
mc = MCEventContainer()
reader = HDF5TableReader(str(temp_h5_file))
for cont in reader.read('/R0/MC', mc):
print(cont)
reader.close()
def test_with_context_reader(test_h5_file):
with HDF5TableReader(test_h5_file) as h5_table:
assert h5_table._h5file.isopen == 1
for cont in h5_table.read("/R0/sim_shower", SimulatedShowerContainer):
print(cont)
assert h5_table._h5file.isopen == 0
def test_write_to_any_location(tmp_path):
path = tmp_path / "test.h5"
loc = "path/path_1"
class ContainerA(Container):
a = Field(0, "some integer field")
container = ContainerA(a=1)
with HDF5TableWriter(path, group_name=loc + "/group_1") as h5:
for _ in range(5):
h5.write("table", container)
h5.write("deeper/table2", container)
with HDF5TableReader(path) as h5:
for container in h5.read(f"/{loc}/group_1/table", ContainerA):
assert container.a == 1
with HDF5TableReader(path) as h5:
for container in h5.read(f"/{loc}/group_1/deeper/table2", ContainerA):
assert container.a == 1
def test_with_context_reader(temp_h5_file):
mc = MCEventContainer()
with HDF5TableReader(temp_h5_file) as h5_table:
assert h5_table._h5file.isopen == 1
for cont in h5_table.read("/R0/MC", mc):
print(cont)
assert h5_table._h5file.isopen == 0
def test_with_context_reader(temp_h5_file):
mc = MCEventContainer()
with HDF5TableReader(str(temp_h5_file)) as h5_table:
assert h5_table._h5file.isopen == True
for cont in h5_table.read('/R0/MC', mc):
print(cont)
assert h5_table._h5file.isopen == False
def test_prefix(tmp_path):
tmp_file = tmp_path / "test_prefix.hdf5"
hillas_parameter_container = HillasParametersContainer(x=1 * u.m,
y=1 * u.m,
length=1 * u.m,
width=1 * u.m)
leakage_container = LeakageContainer(
pixels_width_1=0.1,
pixels_width_2=0.1,
intensity_width_1=0.1,
intensity_width_2=0.1,
)
with HDF5TableWriter(tmp_file.name, group_name="blabla",
add_prefix=True) as writer:
writer.write("events", [hillas_parameter_container, leakage_container])
df = pd.read_hdf(tmp_file.name, key="/blabla/events")
assert "hillas_x" in df.columns
assert "leakage_pixels_width_1" in df.columns
with HDF5TableReader(tmp_file.name) as reader:
generator = reader.read("/blabla/events",
HillasParametersContainer(),
prefix=True)
hillas = next(generator)
for value, read_value in zip(hillas_parameter_container.as_dict().values(),
hillas.as_dict().values()):
np.testing.assert_equal(value, read_value)
with HDF5TableReader(tmp_file.name) as reader:
generator = reader.read("/blabla/events",
LeakageContainer(),
prefix=True)
leakage = next(generator)
for value, read_value in zip(leakage_container.as_dict().values(),
leakage.as_dict().values()):
np.testing.assert_equal(value, read_value)
def test_strings(tmp_path):
"""Test we can write unicode strings"""
from ctapipe.core import Container
from ctapipe.io import read_table
# when not giving a max_len, should be taken from the first container
class Container1(Container):
container_prefix = ""
string = Field("", "test string")
path = tmp_path / "test.h5"
strings = ["Hello", "öäα"]
with HDF5TableWriter(path, mode="w") as writer:
for string in strings:
writer.write("strings", Container1(string=string))
table = read_table(path, "/strings")
# the α is above the max length estimated from the first element
assert table["string"].tolist() == ["Hello", "öä"]
class Container2(Container):
container_prefix = ""
string = Field("", "test string", max_length=10)
path = tmp_path / "test.h5"
strings = ["Hello", "öäα", "12345678910"]
expected = ["Hello", "öäα", "1234567891"]
with HDF5TableWriter(path, mode="w") as writer:
for string in strings:
writer.write("strings", Container2(string=string))
table = read_table(path, "/strings")
# the α is above the max length estimated from the first element
assert table["string"].tolist() == expected
# test this also works with table reader
with HDF5TableReader(path) as reader:
generator = reader.read("/strings", Container2)
for string in expected:
c = next(generator)
assert c.string == string
def test_write_large_integer():
class C(Container):
value = Field(True, "Integer value")
exps = [15, 31, 63]
with tempfile.NamedTemporaryFile() as f:
with HDF5TableWriter(f.name, "test") as writer:
for exp in exps:
c = C(value=2**exp - 1)
writer.write("c", c)
c = C()
with HDF5TableReader(f.name) as reader:
c_reader = reader.read("/test/c", c)
for exp in exps:
cur = next(c_reader)
assert cur.value == 2**exp - 1
def test_write_bool():
class C(Container):
boolean = Field(True, "Boolean value")
with tempfile.NamedTemporaryFile() as f:
with HDF5TableWriter(f.name, "test") as writer:
for i in range(2):
c = C(boolean=(i % 2 == 0))
writer.write("c", c)
c = C()
with HDF5TableReader(f.name) as reader:
c_reader = reader.read("/test/c", c)
for i in range(2):
cur = next(c_reader)
expected = (i % 2) == 0
assert isinstance(cur.boolean, np.bool_)
assert cur.boolean == expected
def test_write_large_integer(tmp_path):
path = tmp_path / "test.h5"
class C(Container):
value = Field(True, "Integer value")
exps = [15, 31, 63]
with HDF5TableWriter(path, "test") as writer:
for exp in exps:
c = C(value=2**exp - 1)
writer.write("c", c)
c = C()
with HDF5TableReader(path) as reader:
c_reader = reader.read("/test/c", c)
for exp in exps:
cur = next(c_reader)
assert cur.value == 2**exp - 1
def test_read_write_container_with_int_enum(tmp_path):
tmp_file = tmp_path / "container_with_int_enum.hdf5"
def create_stream(n_event):
data = WithIntEnum()
for i in range(n_event):
data.event_type = data.EventType(i % 3 + 1)
yield data
with HDF5TableWriter(tmp_file, group_name="data") as h5_table:
for data in create_stream(10):
h5_table.write("table", data)
with HDF5TableReader(tmp_file, mode="r") as h5_table:
for group_name in ["data/"]:
group_name = "/{}table".format(group_name)
for data in h5_table.read(group_name, WithIntEnum()):
assert isinstance(data.event_type, WithIntEnum.EventType)
