def test_divide_offset(datadir, randomdata, fs):
blocks.divide(randomdata, datadir, 10, extension='.pkl')
blocks.divide(randomdata, datadir, 10, 10, extension='.pkl')
assert(len(fs.ls(datadir)) == 20)
df = blocks.assemble(datadir)
expected = randomdata.append(randomdata)
assert(np.isclose(df, expected).all().all())
def test_divide_offset(datadir, randomdata, fs):
blocks.divide(randomdata, datadir, 10, extension=".csv", filesystem=fs)
blocks.divide(randomdata, datadir, 10, 10, extension=".csv", filesystem=fs)
assert len(fs.ls(datadir)) == 20
df = blocks.assemble(datadir)
expected = randomdata.append(randomdata)
assert np.isclose(df, expected).all().all()
def test_divide_cgroups(datadir, randomdata, fs):
randomdata.insert(0, 'key', list(range(10)))
cgroups_columns = {
'cgroup1': ['key', 'f0', 'f1', 'f2'],
'cgroup2': ['key', 'f3', 'f4', 'f5'],
'cgroup3': ['key', 'f6', 'f7', 'f8', 'f9'],
}
blocks.divide(
randomdata,
datadir,
10,
cgroup_columns=cgroups_columns,
extension='.pkl'
)
assert(len(fs.ls(datadir)) == 3)
for subdir in fs.ls(datadir):
assert(len(fs.ls(subdir)) == 10)
df = blocks.assemble(datadir)
assert(df.equals(randomdata))
def test_divide_cgroups(datadir, randomdata, fs):
randomdata.insert(0, "key", list(range(10)))
cgroups_columns = {
"cgroup1": ["key", "f0", "f1", "f2"],
"cgroup2": ["key", "f3", "f4", "f5"],
"cgroup3": ["key", "f6", "f7", "f8", "f9"],
}
blocks.divide(
randomdata,
datadir,
10,
cgroup_columns=cgroups_columns,
extension=".csv",
filesystem=fs,
)
assert len(fs.ls(datadir)) == 3
for subdir in fs.ls(datadir):
assert len(fs.ls(subdir)) == 10
df = blocks.assemble(datadir)
assert np.isclose(df, randomdata).all().all()
def test_divide(datadir, randomdata, fs):
blocks.divide(randomdata, datadir, 10, extension='.pkl')
assert(len(fs.ls(datadir)) == 10)
df = blocks.assemble(datadir)
assert(np.isclose(df, randomdata).all().all())
def test_divide(datadir, randomdata, fs):
blocks.divide(randomdata, datadir, 10, extension=".csv", filesystem=fs)
assert len(fs.ls(datadir)) == 10
df = blocks.assemble(datadir)
assert np.isclose(df, randomdata).all().all()