def test_readwrite_zarr(typ, tmp_path):
X = typ(X_list)
adata_src = ad.AnnData(X, obs=obs_dict, var=var_dict, uns=uns_dict)
adata_src.raw = adata_src
assert not is_categorical_dtype(adata_src.obs["oanno1"])
adata_src.write_zarr(tmp_path / "test_zarr_dir", chunks=True)
adata = ad.read_zarr(tmp_path / "test_zarr_dir")
assert is_categorical_dtype(adata.obs["oanno1"])
assert not is_categorical_dtype(adata.obs["oanno2"])
assert adata.obs.index.tolist() == ["name1", "name2", "name3"]
assert adata.obs["oanno1"].cat.categories.tolist() == ["cat1", "cat2"]
assert adata.obs["oanno1c"].cat.categories.tolist() == ["cat1"]
assert is_categorical_dtype(adata.raw.var["vanno2"])
pd.testing.assert_frame_equal(adata.obs, adata_src.obs)
pd.testing.assert_frame_equal(adata.var, adata_src.var)
assert np.all(adata.var.index == adata_src.var.index)
assert adata.var.index.dtype == adata_src.var.index.dtype
assert type(adata.raw.X) is type(adata_src.raw.X)
assert np.allclose(asarray(adata.raw.X), asarray(adata_src.raw.X))
assert np.all(adata.raw.var == adata_src.raw.var)
assert isinstance(adata.uns["uns4"]["a"], (int, np.integer))
assert isinstance(adata_src.uns["uns4"]["a"], (int, np.integer))
assert type(adata.uns["uns4"]["c"]) is type(adata_src.uns["uns4"]["c"])
assert_equal(adata, adata_src)
def test_inplace_subset_obs(matrix_type, subset_func):
orig = gen_adata((30, 30), X_type=matrix_type)
subset_idx = subset_func(orig.obs_names)
modified = orig.copy()
from_view = orig[subset_idx, :].copy()
modified._inplace_subset_obs(subset_idx)
assert_equal(asarray(from_view.X), asarray(modified.X), exact=True)
assert_equal(from_view.obs, modified.obs, exact=True)
assert_equal(from_view.var, modified.var, exact=True)
for k in from_view.obsm:
assert_equal(asarray(from_view.obsm[k]),
asarray(modified.obsm[k]),
exact=True)
for k in from_view.varm:
assert_equal(asarray(from_view.varm[k]),
asarray(modified.varm[k]),
exact=True)
assert_equal(asarray(orig.varm[k]),
asarray(modified.varm[k]),
exact=True)
for k in from_view.layers:
assert_equal(asarray(from_view.layers[k]),
asarray(modified.layers[k]),
exact=True)
def test_readwrite_h5ad(typ, dataset_kwargs, backing_h5ad):
tmpdir = tempfile.TemporaryDirectory()
tmpdirpth = Path(tmpdir.name)
mid_pth = tmpdirpth / "mid.h5ad"
X = typ(X_list)
adata_src = ad.AnnData(X, obs=obs_dict, var=var_dict, uns=uns_dict)
assert not is_categorical_dtype(adata_src.obs["oanno1"])
adata_src.raw = adata_src
adata_src.write(backing_h5ad, **dataset_kwargs)
adata_mid = ad.read(backing_h5ad)
adata_mid.write(mid_pth, **dataset_kwargs)
adata = ad.read_h5ad(mid_pth)
assert is_categorical_dtype(adata.obs["oanno1"])
assert not is_categorical_dtype(adata.obs["oanno2"])
assert adata.obs.index.tolist() == ["name1", "name2", "name3"]
assert adata.obs["oanno1"].cat.categories.tolist() == ["cat1", "cat2"]
assert is_categorical_dtype(adata.raw.var["vanno2"])
assert np.all(adata.obs == adata_src.obs)
assert np.all(adata.var == adata_src.var)
assert np.all(adata.var.index == adata_src.var.index)
assert adata.var.index.dtype == adata_src.var.index.dtype
assert type(adata.raw.X) is type(adata_src.raw.X)
assert type(adata.raw.varm) is type(adata_src.raw.varm)
assert np.allclose(asarray(adata.raw.X), asarray(adata_src.raw.X))
assert np.all(adata.raw.var == adata_src.raw.var)
assert isinstance(adata.uns["uns4"]["a"], (int, np.integer))
assert isinstance(adata_src.uns["uns4"]["a"], (int, np.integer))
assert type(adata.uns["uns4"]["c"]) is type(adata_src.uns["uns4"]["c"])
assert_equal(adata, adata_src)
def test_double_index(subset_func, subset_func2):
adata = gen_adata((10, 10))
obs_subset = subset_func(adata.obs_names)
var_subset = subset_func2(adata.var_names)
v1 = adata[obs_subset, var_subset]
v2 = adata[obs_subset, :][:, var_subset]
assert np.all(asarray(v1.X) == asarray(v2.X))
assert np.all(v1.obs == v2.obs)
assert np.all(v1.var == v2.var)
def test_view_of_view_modification():
adata = ad.AnnData(np.zeros((10, 10)))
adata[0, :][:, 5:].X = np.ones(5)
assert np.all(adata.X[0, 5:] == np.ones(5))
adata[[1, 2], :][:, [1, 2]].X = np.ones((2, 2))
assert np.all(adata.X[1:3, 1:3] == np.ones((2, 2)))
adata.X = sparse.csr_matrix(adata.X)
adata[0, :][:, 5:].X = np.ones(5) * 2
assert np.all(asarray(adata.X)[0, 5:] == np.ones(5) * 2)
adata[[1, 2], :][:, [1, 2]].X = np.ones((2, 2)) * 2
assert np.all(asarray(adata.X)[1:3, 1:3] == np.ones((2, 2)) * 2)
def test_set_scalar_subset_X(matrix_type, subset_func):
adata = ad.AnnData(matrix_type(np.zeros((10, 10))))
orig_X_val = adata.X.copy()
subset_idx = slice_subset(adata.obs_names)
adata_subset = adata[subset_idx, :]
adata_subset.X = 1
assert adata_subset.is_view
assert np.all(asarray(adata[subset_idx, :].X) == 1)
assert asarray((orig_X_val != adata.X)).sum() == mul(*adata_subset.shape)
def test_read_write_X(tmp_path, mtx_format, backed_mode, force_dense):
base_pth = Path(tmp_path)
orig_pth = base_pth / "orig.h5ad"
backed_pth = base_pth / "backed.h5ad"
orig = ad.AnnData(mtx_format(asarray(sparse.random(10, 10, format="csr"))))
orig.write(orig_pth)
backed = ad.read(orig_pth, backed=backed_mode)
backed.write(backed_pth, as_dense=["X"])
backed.file.close()
from_backed = ad.read(backed_pth)
assert np.all(asarray(orig.X) == asarray(from_backed.X))
def test_assigmnent_dict(adata):
d_obsp = dict(
a=pd.DataFrame(np.ones((M, M)), columns=adata.obs_names, index=adata.obs_names),
b=np.zeros((M, M)),
c=sparse.random(M, M, format="csr"),
)
d_varp = dict(
a=pd.DataFrame(np.ones((N, N)), columns=adata.var_names, index=adata.var_names),
b=np.zeros((N, N)),
c=sparse.random(N, N, format="csr"),
)
adata.obsp = d_obsp
for k, v in d_obsp.items():
assert np.all(asarray(adata.obsp[k]) == asarray(v))
adata.varp = d_varp
for k, v in d_varp.items():
assert np.all(asarray(adata.varp[k]) == asarray(v))
def assert_equal_ndarray(a, b, exact=False, elem_name=None):
b = asarray(b)
if not exact and is_numeric_dtype(a) and is_numeric_dtype(b):
assert a.shape == b.shape, format_msg(elem_name)
assert np.allclose(a, b, equal_nan=True), format_msg(elem_name)
elif ( # Structured dtype
not exact and hasattr(a, "dtype") and hasattr(b, "dtype")
and len(a.dtype) > 1 and len(b.dtype) > 0):
assert_equal(pd.DataFrame(a), pd.DataFrame(b), exact, elem_name)
else:
assert np.all(a == b), format_msg(elem_name)
def test_not_set_subset_X(matrix_type, subset_func):
adata = ad.AnnData(matrix_type(asarray(sparse.random(20, 20))))
init_hash = joblib.hash(adata)
orig_X_val = adata.X.copy()
while True:
subset_idx = slice_subset(adata.obs_names)
if len(adata[subset_idx, :]) > 2:
break
subset = adata[subset_idx, :]
subset = adata[:, subset_idx]
internal_idx = _normalize_index(subset_func(np.arange(subset.X.shape[1])),
subset.var_names)
assert subset.is_view
subset.X[:, internal_idx] = 1
assert not subset.is_view
assert not np.any(asarray(adata.X != orig_X_val))
assert init_hash == joblib.hash(adata)
def test_inplace_subset_var(matrix_type, subset_func):
orig = gen_adata((30, 30), X_type=matrix_type)
subset_idx = subset_func(orig.var_names)
modified = orig.copy()
from_view = orig[:, subset_idx].copy()
modified._inplace_subset_var(subset_idx)
assert_array_equal(asarray(from_view.X), asarray(modified.X))
assert_array_equal(from_view.obs, modified.obs)
assert_array_equal(from_view.var, modified.var)
for k in from_view.obsm:
assert_array_equal(asarray(from_view.obsm[k]),
asarray(modified.obsm[k]))
assert_array_equal(asarray(orig.obsm[k]), asarray(modified.obsm[k]))
for k in from_view.varm:
assert_array_equal(asarray(from_view.varm[k]),
asarray(modified.varm[k]))
for k in from_view.layers:
assert_array_equal(asarray(from_view.layers[k]),
asarray(modified.layers[k]))
def test_concatenate_layers_outer(array_type, fill_val):
# Testing that issue #368 is fixed
a = AnnData(
X=np.ones((10, 20)),
layers={"a": array_type(sparse.random(10, 20, format="csr"))},
)
b = AnnData(X=np.ones((10, 20)))
c = a.concatenate(b, join="outer", fill_value=fill_val, batch_categories=["a", "b"])
np.testing.assert_array_equal(
asarray(c[c.obs["batch"] == "b"].layers["a"]), fill_val
)
def test_modify_view_component(matrix_type, mapping_name):
adata = ad.AnnData(
np.zeros((10, 10)),
**{mapping_name: dict(m=matrix_type(asarray(sparse.random(10, 10))))},
)
init_hash = joblib.hash(adata)
subset = adata[:5, :][:, :5]
assert subset.is_view
m = getattr(subset, mapping_name)["m"]
m[0, 0] = 100
assert not subset.is_view
assert getattr(subset, mapping_name)["m"][0, 0] == 100
assert init_hash == joblib.hash(adata)
def test_concatenate_fill_value(fill_val):
def get_obs_els(adata):
return {
"X": adata.X,
**{f"layer_{k}": adata.layers[k]
for k in adata.layers},
**{f"obsm_{k}": adata.obsm[k]
for k in adata.obsm},
}
adata1 = gen_adata((10, 10))
adata1.obsm = {
k: v
for k, v in adata1.obsm.items() if not isinstance(v, pd.DataFrame)
}
adata2 = gen_adata((10, 5))
adata2.obsm = {
k: v[:, :v.shape[1] // 2]
for k, v in adata2.obsm.items() if not isinstance(v, pd.DataFrame)
}
adata3 = gen_adata((7, 3))
adata3.obsm = {
k: v[:, :v.shape[1] // 3]
for k, v in adata3.obsm.items() if not isinstance(v, pd.DataFrame)
}
joined = adata1.concatenate([adata2, adata3],
join="outer",
fill_value=fill_val)
ptr = 0
for orig in [adata1, adata2, adata3]:
cur = joined[ptr:ptr + orig.n_obs]
cur_els = get_obs_els(cur)
orig_els = get_obs_els(orig)
for k, cur_v in cur_els.items():
orig_v = orig_els.get(k, sparse.csr_matrix((orig.n_obs, 0)))
assert_equal(cur_v[:, :orig_v.shape[1]], orig_v)
np.testing.assert_equal(asarray(cur_v[:, orig_v.shape[1]:]),
fill_val)
ptr += orig.n_obs
def assert_equal_h5py_dataset(a, b, exact=False, elem_name=None):
a = asarray(a)
assert_equal(b, a, exact, elem_name=elem_name)
def assert_equal_sparse(a, b, exact=False, elem_name=None):
a = asarray(a)
assert_equal(b, a, exact, elem_name=elem_name)
def assert_equal_arrayview(a, b, exact=False, elem_name=None):
assert_equal(asarray(a), asarray(b), exact=exact, elem_name=elem_name)
def test_setter_singular_dim(shape, orig_array_type, new_array_type):
# https://github.com/theislab/anndata/issues/500
adata = gen_adata(shape, X_type=orig_array_type)
adata.X = new_array_type(np.ones(shape))
np.testing.assert_equal(asarray(adata.X), 1)