def pytest_generate_tests(metafunc):
_test_suite = get_series()
if metafunc.function.__name__ == "test_relations":
_series_map = get_convert_map()
all_relations_tested(_series_map)
argsvalues = []
for item in _test_suite:
for source_type, relation_type, series_list in _series_map:
if item in relation_type:
args = {
"id": f"{item.name}: {relation_type} -> {source_type}"
}
if item.name not in series_list:
args["marks"] = pytest.mark.xfail(raises=ValueError)
argsvalues.append(
pytest.param(source_type, relation_type, item, **args))
metafunc.parametrize(
argnames=["source_type", "relation_type", "series"],
argvalues=argsvalues)
if metafunc.function.__name__ in [
"test_consistency",
"test_side_effects",
"test_multiple_inference",
]:
argsvalues = []
for series in _test_suite:
args = {"id": f"{series.name}"}
argsvalues.append(pytest.param(series, **args))
metafunc.parametrize(argnames=["series"], argvalues=argsvalues)
def pytest_generate_tests(metafunc):
_test_suite = get_series()
if metafunc.function.__name__ in [
"test_consistency", "test_traversal_mutex"
]:
argsvalues = []
for series in _test_suite:
args = {"id": series.name}
argsvalues.append(pytest.param(series, **args))
metafunc.parametrize(argnames=["series"], argvalues=argsvalues)
if metafunc.function.__name__ == "test_inference":
argsvalues = []
inferred_series_type_map = infer_series_type_map()
for series in _test_suite:
expected_type = inferred_series_type_map[series.name]
for test_type in typeset.types:
args = {
"id":
"{name} x {type} expected {expected}".format(
name=series.name,
type=test_type,
expected=test_type == expected_type,
)
}
difference = test_type != expected_type
argsvalues.append(
pytest.param(series, test_type, typeset, difference,
**args))
metafunc.parametrize(
argnames=["series", "expected_type", "typeset", "difference"],
argvalues=argsvalues,
)
def relations_transform_test():
convert_map = get_convert_map()
relation_tests = {
get_relation(*conversions[0:2]): conversions[2]
for conversions in convert_map
}
series_dict = {s.name: s for s in get_series()}
performance_list = []
for relation, names in relation_tests.items():
test_series = {name: series_dict[name] for name in names}
performance_list.extend(
profile_relation_transform(relation, test_series))
df = pd.DataFrame.from_records(performance_list)
grouper = "relation"
df[grouper] = df[grouper].astype(str)
aggs = ["min", "max"]
agg_labels = ["worst", "best"]
summary_cols = ["series", "big O"]
agg_df = df.groupby(grouper).agg(aggs).reset_index()[[grouper] +
summary_cols]
agg_df.columns = ["_".join(col).strip("_") for col in agg_df.columns]
colrenames = {
f"{name}_{agg}": f"{rename} {name}"
for name in summary_cols for agg, rename in zip(aggs, agg_labels)
}
agg_df.rename(columns=colrenames, inplace=True)
df["normed run time"] = df["average run time"] / df[
"average run time"].min()
df = df.groupby(grouper)["normed run time"].describe().sort_values("50%")
df = pd.merge(df, agg_df, on=grouper, how="left")
return df
def pytest_generate_tests(metafunc):
_test_suite = get_series()
if metafunc.function.__name__ in [
"test_consistency", "test_traversal_mutex"
]:
argsvalues = []
for series in _test_suite:
args = {"id": series.name}
argsvalues.append(pytest.param(series, **args))
metafunc.parametrize(argnames=["series"], argvalues=argsvalues)
def pytest_generate_tests(metafunc):
_test_suite = get_series()
if metafunc.function.__name__ in [
"test_consistency",
"test_side_effects",
"test_multiple_inference",
]:
argsvalues = []
for series in _test_suite:
args = {"id": str(series.name)}
argsvalues.append(pytest.param(series, **args))
metafunc.parametrize(argnames=["series"], argvalues=argsvalues)
def pytest_generate_tests(metafunc):
_test_suite = get_series()
if metafunc.function.__name__ == "test_contains":
_series_map = get_contains_map()
all_series_included(_test_suite, _series_map)
argsvalues = []
for item in _test_suite:
for type, series_list in _series_map.items():
args = {
"id":
"{item_name} x {item_type}".format(item_name=item.name,
item_type=type)
}
member = item.name in series_list
argsvalues.append(pytest.param(item, type, member, **args))
metafunc.parametrize(argnames=["series", "type", "member"],
argvalues=argsvalues)
def pytest_generate_tests(metafunc):
_test_suite = get_series()
if metafunc.function.__name__ == "test_contains":
_series_map = get_contains_map()
all_series_included(_test_suite, _series_map)
argsvalues = []
for item in _test_suite:
for type, series_list in _series_map.items():
args = {
"id":
"{item_name} x {item_type}".format(item_name=item.name,
item_type=type)
}
if item.name not in series_list:
args["marks"] = pytest.mark.xfail(raises=AssertionError)
argsvalues.append(pytest.param(item, type, **args))
metafunc.parametrize(argnames=["series", "type"], argvalues=argsvalues)
def pytest_generate_tests(metafunc):
_test_suite = get_series()
if metafunc.function.__name__ == "test_relations":
_series_map = get_convert_map()
all_relations_tested(_series_map)
argsvalues = []
for item in _test_suite:
for source_type, relation_type, series_list in _series_map:
if item in relation_type:
args = {
"id":
"{name}: {relation_type} -> {source_type}".format(
name=item.name,
relation_type=relation_type,
source_type=source_type,
)
}
member = item.name in series_list
argsvalues.append(
pytest.param(source_type, relation_type, item, member,
**args))
metafunc.parametrize(
argnames=["source_type", "relation_type", "series", "member"],
argvalues=argsvalues,
)
if metafunc.function.__name__ in [
"test_consistency",
"test_side_effects",
"test_multiple_inference",
]:
argsvalues = []
for series in _test_suite:
args = {"id": str(series.name)}
argsvalues.append(pytest.param(series, **args))
metafunc.parametrize(argnames=["series"], argvalues=argsvalues)
def performance_report(membership=True):
series_dict = {s.name: s for s in get_series()}
conv_map = get_contains_map()
performance_list = []
for type, series_names in conv_map.items():
if membership:
# True: "series in type"
test_series = {name: series_dict[name] for name in series_names}
else:
# False: "series in type"
test_series = {
s.name: s
for s in series_dict.values() if s.name not in series_names
}
performance_list.extend(profile_type(type, test_series))
df = pd.DataFrame.from_records(performance_list)
df["type"] = df["type"].astype(str)
aggs = ["min", "max"]
agg_labels = ["worst", "best"]
summary_cols = ["series", "big O"]
agg_df = df.groupby("type").agg(aggs).reset_index()[["type"] +
summary_cols]
agg_df.columns = ["_".join(col).strip("_") for col in agg_df.columns]
colrenames = {
f"{name}_{agg}": f"{rename} {name}"
for name in summary_cols for agg, rename in zip(aggs, agg_labels)
}
agg_df.rename(columns=colrenames, inplace=True)
df["normed run time"] = df["average run time"] / df[
"average run time"].min()
df = df.groupby("type")["normed run time"].describe().sort_values("50%")
df = pd.merge(df, agg_df, on="type", how="left")
return df
File,
Float,
Generic,
Geometry,
Image,
Integer,
IPAddress,
Object,
Ordinal,
Path,
String,
Time,
TimeDelta,
)
series = get_series()
typeset = CompleteSet()
contains_map = {
Integer: {
"int_series",
"Int64_int_series",
"int_range",
"Int64_int_nan_series",
"int_series_boolean",
},
Count: {"np_uint32"},
Path:
{"path_series_linux", "path_series_linux_missing", "path_series_windows"},
URL: {"url_series", "url_nan_series", "url_none_series"},
