def test_unzip_and_convert_metadata(self):
file_zip = os.path.join(TestAsvJsonText.data, 'results2.zip')
temp = get_temp_folder(__file__, 'temp_unzip_and_convert_metadata')
create_asv_benchmark(location=temp,
models={'LogisticRegression', 'LinearRegression'})
unzip_files(file_zip, temp)
data = os.path.join(temp, 'results')
conf = os.path.join(temp, 'asv.conf.json')
exp = export_asv_json(data, baseline="skl", conf=conf)
par_problem = []
par_scenario = []
for row in exp:
if 'par_problem' in row:
par_problem.append(row['par_problem'])
if 'par_scenario' in row:
par_scenario.append(row['par_scenario'])
s = set(par_scenario)
self.assertEqual(s, {'default', 'liblinear'})
s = set(par_problem)
self.assertEqual(s, {
'm-cl', '~m-reg-64', 'b-cl', 'm-reg', 'b-reg', '~b-cl-64',
'~b-reg-64'
})
out = os.path.join(temp, "df.xlsx")
df = pandas.DataFrame(exp)
df.to_excel(out)
def test_cli_asv2csv(self):
temp = get_temp_folder(__file__, "temp_asv2csv")
file_zip = os.path.join(TestCliAsvBench.data, 'results.zip')
unzip_files(file_zip, temp)
data = os.path.join(temp, 'results')
out = os.path.join(temp, "data.csv")
st = BufferedPrint()
main(args=["asv2csv", "-f", data, "-o", out], fLOG=st.fprint)
self.assertExists(out)
df = pandas.read_csv(out)
self.assertEqual(df.shape, (168, 66))
out = os.path.join(temp, "data<date>.csv")
main(args=["asv2csv", "-f", data, "-o", out], fLOG=st.fprint)
def test_unzip_and_convert(self):
file_zip = os.path.join(TestAsvJsonText.data, 'results.zip')
temp = get_temp_folder(__file__, 'temp_unzip_and_convert')
unzip_files(file_zip, temp)
data = os.path.join(temp, 'results')
exp = export_asv_json(data, baseline="skl")
self.assertIsInstance(exp, list)
self.assertTrue(all(map(lambda x: isinstance(x, dict), exp)))
cc = 0
for e in exp:
ms = [k for k in e if k.startswith("M-")]
rs = [k for k in e if k.startswith("R-")]
if len(ms) > 0 and len(rs) > 0:
cc += 1
if cc == 0:
raise AssertionError("No rs")
def test_process_data(self):
fLOG(
__file__,
self._testMethodName,
OutputPrint=__name__ == "__main__")
if is_travis_or_appveyor():
warnings.warn("disabled on appveyor and travis")
return
temp = get_temp_folder(__file__, "temp_process_data_cresus_2016")
import keyring
pwd = keyring.get_password(
"cresus", os.environ["COMPUTERNAME"] + "ensae")
assert pwd
name = cresus_dummy_file()
if not os.path.exists(name):
raise FileNotFoundError(name)
zipname = os.path.join(temp, "bdd.zip")
pwd = pwd.encode("ascii")
decrypt_stream(pwd, name, zipname)
res = unzip_files(zipname, temp)
fLOG(res)
infile = res[0]
train, test = process_cresus_whole_process(
infile, outfold=temp, fLOG=fLOG)
for r in train.values():
df = pandas.read_csv(r, sep="\t", encoding="utf-8")
fLOG(df.columns)