def test_histogramdd_arrays_and_series():
x = da.random.standard_normal(size=(1000, ), chunks=(200, ))
y = da.random.standard_normal(size=(1000, ),
chunks=(200, )).to_dask_dataframe()
w = da.random.uniform(size=(1000, ), chunks=(200, ))
bins = (12, 12)
range = ((-3, 3), (-2.6, 2.6))
h1 = dhb.histogramdd(
(x, y),
bins=bins,
range=range,
weights=w,
histogram=dhb.Histogram,
storage=dhb.storage.Weight(),
)
h1.compute()
h2 = bhnp.histogramdd(
(x.compute(), y.compute()),
bins=bins,
range=range,
weights=w.compute(),
histogram=bh.Histogram,
storage=bh.storage.Weight(),
)
np.testing.assert_array_almost_equal(h1.view()["value"],
h2.view()["value"])
np.testing.assert_array_almost_equal(h1.view()["variance"],
h2.view()["variance"])
def test_histogramdd_multicolumn_input():
x = da.random.standard_normal(size=(3_000, 3), chunks=(500, 3))
bins = [[-3, -2, 2, 3], [-4, -1, 1, 4], [-2, -1, 1, 2]]
h1 = dhb.histogramdd(x, bins=bins, histogram=dhb.Histogram)
h2 = bhnp.histogramdd(x.compute(), bins=bins, histogram=bh.Histogram)
h1 = h1.compute()
np.testing.assert_array_almost_equal(h1.view(), h2.view())
def test_histogramdd():
x = da.random.standard_normal(size=(3_000, ), chunks=500)
y = da.random.standard_normal(size=(3_000, ), chunks=500)
z = da.random.standard_normal(size=(3_000, ), chunks=500)
w = da.random.uniform(0.1, 0.5, size=(3_000, ), chunks=500)
bins = (4, 5, 6)
range = ((-2.5, 2.5), (-3.5, 3.5), (-2, 2))
h1 = dhb.histogramdd(
(x, y, z),
bins=bins,
range=range,
weights=w,
histogram=dhb.Histogram,
storage=dhb.storage.Weight(),
)
h2 = bhnp.histogramdd(
(x.compute(), y.compute(), z.compute()),
bins=bins,
range=range,
weights=w,
histogram=bh.Histogram,
storage=bh.storage.Weight(),
)
h1 = h1.compute()
np.testing.assert_array_almost_equal(h1.view()["value"],
h2.view()["value"])
np.testing.assert_array_almost_equal(h1.view()["variance"],
h2.view()["variance"])
def test_histogramdd_da_return():
x = da.random.standard_normal(size=(3_000, ), chunks=500)
y = da.random.standard_normal(size=(3_000, ), chunks=500)
z = da.random.standard_normal(size=(3_000, ), chunks=500)
w = da.random.uniform(0.1, 0.5, size=(3_000, ), chunks=500)
bins = (4, 5, 6)
range = ((-2.5, 2.5), (-3.5, 3.5), (-2, 2))
h1, edges1 = dhb.histogramdd((x, y, z), bins=bins, range=range, weights=w)
h2, edges2 = bhnp.histogramdd((x.compute(), y.compute(), z.compute()),
bins=bins,
range=range,
weights=w)
np.testing.assert_array_almost_equal(h1.compute(), h2)
for e1, e2 in zip(edges1, edges2):
np.testing.assert_array_almost_equal(e1.compute(), e2)
def test_histogram_da_return():
x = da.random.standard_normal(size=(3_000, ), chunks=500)
bins = np.array([-3, -2.2, 0, 1.1, 2.2, 3.3])
h1, edges1 = dhb.histogram(x, bins=bins)
h2, edges2 = bhnp.histogram(x.compute(), bins=bins)
np.testing.assert_array_almost_equal(h1.compute(), h2)
np.testing.assert_array_almost_equal(edges1.compute(), edges2)
x = da.random.standard_normal(size=(3_000, 3), chunks=(500, 3))
bins = (5, 6, 7)
range = ((-3, 3), ) * 3
h1, edges1 = dhb.histogramdd(x, bins=bins, range=range)
h2, edges2 = bhnp.histogramdd(x.compute(), bins=bins, range=range)
np.testing.assert_array_almost_equal(h1.compute(), h2)
for e1, e2 in zip(edges1, edges2):
np.testing.assert_array_almost_equal(e1.compute(), e2)
def test_histogramdd_dataframe():
x = da.random.standard_normal(size=(1000, 3), chunks=(200, 3))
df = x.to_dask_dataframe(columns=["a", "b", "c"])
w = da.random.uniform(size=(1000, ), chunks=(200, ))
bins = [[-3, -2, 1, 2, 3], [-4, -1, 1, 2, 3, 4], [-3, -2, 0, 2]]
h1 = dhb.histogramdd(
df,
bins=bins,
weights=w,
histogram=dhb.Histogram,
storage=dhb.storage.Weight(),
)
h1.compute()
h2 = bhnp.histogramdd(
df.compute().to_numpy(),
bins=bins,
weights=w.compute(),
histogram=bh.Histogram,
storage=bh.storage.Weight(),
)
np.testing.assert_array_almost_equal(h1.view()["value"],
h2.view()["value"])
np.testing.assert_array_almost_equal(h1.view()["variance"],
h2.view()["variance"])