def test_vanilla_get_expression_data(testfiles):
out = allen.get_expression_data(testfiles, ATLAS.image)
assert isinstance(out, pd.DataFrame)
assert out.index.name == 'label'
assert out.columns.name == 'gene_symbol'
with pytest.raises(KeyError):
allen.get_expression_data({'microarray': [1, 2, 3]}, ATLAS.image)
def test_get_expression_data_errors(testfiles, atlas):
# invalid probe_selection method
with pytest.raises(ValueError):
allen.get_expression_data(atlas['image'], donors=['12876', '15496'],
probe_selection='nonsense')
# cannot use diff_stability with only one donor
with pytest.raises(ValueError):
allen.get_expression_data(atlas['image'], donors=['12876'],
probe_selection='diff_stability')
def test_get_expression_data_errors(testfiles, atlas):
# invalid probe_selection method
with pytest.raises(ValueError):
allen.get_expression_data(atlas['image'],
donors=testfiles.keys(),
probe_selection='nonsense')
# cannot use diff_stability with only one donor
with pytest.raises(ValueError):
donor = list(testfiles.keys())[0]
allen.get_expression_data(atlas['image'],
donors=[donor],
probe_selection='diff_stability')
def test_missing_labels(testfiles, atlas):
# remove some labels from atlas image so numbers are non-sequential
remove = [10, 20, 60]
# subset atlas image
img = check_img(atlas['image'])
img_data = np.asarray(img.dataobj)
for i in remove:
img_data[img_data == i] = 0
img = img.__class__(img_data, img.affine)
# subset atlas info
info = pd.read_csv(atlas['info'])
info = info[~info['id'].isin(remove)]
# test get expression
out, counts = allen.get_expression_data(img,
info,
exact=False,
return_counts=True,
donors=['12876', '15496'])
assert out.index.name == 'label'
assert out.columns.name == 'gene_symbol'
assert len(out) == len(info)
assert isinstance(counts, pd.DataFrame)
assert counts.shape == (len(info), len(testfiles))
def donor_expression(testdir, testfiles):
return allen.get_expression_data(ATLAS.image,
ATLAS.info,
exact=False,
return_donors=True,
data_dir=testdir,
donors=['12876', '15496'])
def test_extra_get_expression_data(testfiles, atlas, opts):
if opts.get('atlas_info'):
opts['atlas_info'] = atlas['info']
out = allen.get_expression_data(atlas['image'], donors=['12876', '15496'],
**opts)
assert out.index.name == 'label'
assert out.columns.name == 'gene_symbol'
def test_extra_get_expression_data(testfiles):
for opts in [{'atlas_info': ATLAS.info},
{'exact': False},
{'reannotated': False},
{'atlas_info': ATLAS.info, 'exact': False}]:
out = allen.get_expression_data(testfiles, ATLAS.image, **opts)
assert isinstance(out, pd.DataFrame)
assert out.index.name == 'label'
assert out.columns.name == 'gene_symbol'
def test_extra_get_expression_data(testfiles, atlas, opts):
if opts.get('atlas_info', False):
opts['atlas_info'] = atlas['info']
out = allen.get_expression_data(atlas['image'],
donors=testfiles.keys(),
**opts)
assert out.index.name == 'label'
assert out.columns.name == 'gene_symbol'
if 'missing' in opts and 'atlas_info' not in opts:
assert not np.any(out.isna())
elif 'missing' in opts and 'atlas_info' in opts:
assert np.any(out.isna())
def test_missing_labels(testfiles):
# remove some labels from atlas image so numbers are non-sequential
remove = [10, 20, 60]
# subset atlas image
atlas = check_niimg(ATLAS.image).get_data()
for i in remove:
atlas[atlas == i] = 0
atlas = new_img_like(ATLAS.image, atlas)
# subset atlas info
atlas_info = pd.read_csv(ATLAS.info)
atlas_info = atlas_info[~atlas_info.id.isin(remove)]
# test get expression
out, counts = allen.get_expression_data(testfiles, atlas, atlas_info,
exact=False, return_counts=True)
assert isinstance(out, pd.DataFrame)
assert out.index.name == 'label'
assert out.columns.name == 'gene_symbol'
assert len(out) == len(atlas_info)
assert isinstance(counts, pd.DataFrame)
assert counts.shape == (len(atlas_info), len(testfiles.probes))
def test_vanilla_get_expression_data(testfiles, atlas):
out = allen.get_expression_data(atlas['image'], donors=['12876', '15496'])
assert np.allclose(out.index, range(1, 84))
assert out.index.name == 'label'
assert out.columns.name == 'gene_symbol'
def test_individualized_surface_expression_data(testfiles):
atlas = allen.datasets.fetch_desikan_killiany(native=True, surface=True)
out = allen.get_expression_data(atlas['image'], donors=testfiles.keys())
assert out.index.name == 'label'
assert out.columns.name == 'gene_symbol'
def test_vanilla_surface_expression_data(testfiles, surface):
out = allen.get_expression_data(surface['image'], donors=testfiles.keys())
assert out.index.name == 'label'
assert out.columns.name == 'gene_symbol'
def test_vanilla_get_expression_data(testdir, testfiles):
out = allen.get_expression_data(ATLAS.image, data_dir=testdir,
donors=['12876', '15496'])
assert isinstance(out, pd.DataFrame)
assert out.index.name == 'label'
assert out.columns.name == 'gene_symbol'
def donor_expression(testfiles):
return allen.get_expression_data(testfiles, ATLAS.image, ATLAS.info,
exact=False, return_donors=True)
def donor_expression(testfiles, atlas):
return list(
allen.get_expression_data(atlas['image'],
missing='centroids',
return_donors=True,
donors=['12876', '15496']).values())
def donor_expression(testfiles, atlas):
return allen.get_expression_data(atlas['image'],
atlas['info'],
exact=False,
return_donors=True,
donors=['12876', '15496'])
