def base_annotation_test(self,
temp,
length,
dtypes,
ann_dtypes,
head=True,
on="cell_id"):
"""
base test of annotate_csv
:param temp: tempdir to test in
:param length: length of test dfs
:param dtypes: dtypes of test dfs
:param ann_dtypes: dtypes of test annotation dict
:param write_header: T/F write header post-annotation
:param on: col to annotate on:
"""
csv, annotation = self.make_ann_test_inputs(temp,
length,
dtypes,
ann_dtypes,
write_header=head,
on=on)
annotated = os.path.join(temp, "annotated.csv.gz")
csvutils.annotate_csv(csv,
annotation,
annotated,
ann_dtypes,
write_header=head,
on=on)
return csv, annotation, annotated
def annotate_coverage_metrics(metrics, coverage_yaml, output):
data = {}
for cell_id, filename in coverage_yaml.items():
with open(filename, 'rt') as reader:
covdata = yaml.load(reader)
if 'cell_id' in covdata:
assert covdata['cell_id'] == cell_id
del covdata['cell_id']
data[cell_id] = covdata
csvutils.annotate_csv(metrics, data, output, dtypes()['metrics'])
def add_clustering_order(
reads, metrics, output, chromosomes=None, sample_info=None):
"""
adds sample information to metrics in place
"""
order = get_hierarchical_clustering_order(
reads, chromosomes=chromosomes
)
if sample_info:
for cell_id, order in order.items():
sample_info[cell_id]['order'] = order
else:
sample_info = order
csvutils.annotate_csv(metrics, sample_info, output)
def add_clustering_order(
reads, metrics, output, chromosomes=None, sample_info=None):
"""
adds sample information to metrics in place
"""
order = get_hierarchical_clustering_order(
reads, chromosomes=chromosomes
)
if not sample_info:
sample_info = {}
for cell_id, order in order.items():
if cell_id not in sample_info:
sample_info[cell_id] = {}
sample_info[cell_id]['order'] = order
csvutils.annotate_csv(metrics, sample_info, output, dtypes()['metrics'])
def test_annotate_csv_annotation_col_mismatch(self, tmpdir, n_rows):
"""
test annotating csv where annotation_data differs in length from csv
:param tmpdir: temporary directory to write in
:param n_rows: number of rows in test csvs
"""
dtypes = {v: "int" for v in 'ABCD'}
dtypes["cell_id"] = "str"
ann_dtypes = {v: "int" for v in 'ERF'}
annotated = os.path.join(tmpdir, "annotated.csv.gz")
csv, annotation = self.make_ann_test_inputs(tmpdir, n_rows, dtypes,
ann_dtypes)
annotation["new_cell"] = {"E": 1, "R": 43, "F": 2}
csvutils.annotate_csv(csv, annotation, annotated, ann_dtypes)
self.validate_annotation_test(csv, annotation, annotated, "cell_id")
def test_annotate_csv_annotation_col_dtype_mismatch(self, tmpdir, n_rows):
"""
test annotating csv with inappropriate annotation_dtypes
:param tmpdir: temporary directory to write in
:param n_rows: number of rows in test csvs
"""
dtypes = {v: "int" for v in 'ABCD'}
dtypes["cell_id"] = "str"
ann_dtypes = {v: "int" for v in 'ERF'}
annotated = os.path.join(tmpdir, "annotated.csv.gz")
csv, annotation = self.make_ann_test_inputs(tmpdir, n_rows, dtypes,
ann_dtypes)
new_keys = range(len(annotation.keys()))
annotation = {new_keys[i]: annotation[cell_id]
for i, cell_id in enumerate(annotation.keys())}
csvutils.annotate_csv(csv, annotation, annotated, ann_dtypes)
assert self.dfs_exact_match(annotated, csv)