def test_convert_miniml_control():
geo_id = 'GSE99650'
gsm_id = 'GSM2649283'
test_data_dir = f'docs/example_data/{geo_id}' # created by this in test environment
testargs = ["__program__", '-i', geo_id, '-d', test_data_dir, '-k', 'SKTR']
with patch.object(sys, 'argv', testargs):
convert_miniml(
geo_id,
data_dir=test_data_dir,
merge=True,
download_it=True,
extract_controls=True,
require_keyword=None,
sync_idats=True,
remove_tgz=True,
verbose=True)
files_found = list(Path(test_data_dir).rglob('*'))
if len(files_found) != 3:
raise AssertionError("Did not download all the files.")
samplesheet = pd.read_csv(Path(test_data_dir,f'{geo_id}_GPL13534_samplesheet.csv'))
if samplesheet['GSM_ID'][0] != gsm_id:
raise AssertionError("Samplesheet did not contain the right data.")
if len(samplesheet['GSM_ID']) != 1:
raise AssertionError("Control filtering failed")
for file in Path(test_data_dir).rglob('*.xml'):
file.unlink()
for file in Path(test_data_dir).rglob('*.pkl'):
file.unlink()
for file in Path(test_data_dir).rglob('*.csv'):
file.unlink()
for file in Path(test_data_dir).rglob('*.tgz'):
file.unlink()
Path(test_data_dir).rmdir()
def test_convert_miniml():
geo_id = 'GSE17769'
test_data_dir = f'docs/example_data/{geo_id}' # created by this in test environment
testargs = ["__program__", '-i', geo_id, '-d', test_data_dir]
with patch.object(sys, 'argv', testargs):
convert_miniml(geo_id,
data_dir=test_data_dir,
merge=True,
download_it=True,
extract_controls=False,
require_keyword=None,
sync_idats=True,
remove_tgz=True,
verbose=True)
#for file in Path(test_data_dir).rglob('*.xml'):
# file.unlink()
shutil.rmtree(test_data_dir)