def test_DPL():
df = pd.DataFrame({"x": ["a", "a", "b", "b"], "y": [1, 1, 0, 1]})
res = metric_one_vs_all(DPL,
df["x"],
label=df["y"],
positive_label_index=(df["y"] == 1))
assert res["a"] == -0.5
assert res["b"] == 0.5
return
def test_CI():
sensitive_facet_index = dfB[0] == "F"
assert CI(dfB[0], sensitive_facet_index) == approx(-1 / 6)
sensitive_facet_index = dfB[0] == "M"
assert CI(dfB[0], sensitive_facet_index) == approx(1 / 6)
# Continuous Facet, Binary Label
sensitive_facet_index = dfC[0] > 1.0
assert CI(dfC[0], sensitive_facet_index) == approx(-1 / 3)
sensitive_facet_index = dfC[0] < 1.0
assert CI(dfC[0], sensitive_facet_index) == approx(1 / 3)
# Multicategory Facet, Binary Label
response = metric_one_vs_all(CI, dfM[0])
assert response["M"] == approx(1 / 3)
assert response["F"] == approx(1 / 4)
assert response["O"] == approx(5 / 12)
def test_CDD():
x = pd.Series([
"M",
"M",
"M",
"F",
"F",
"F",
"F",
"M",
"M",
"M",
"M",
"F",
"M",
"M",
"F",
"M",
"F",
"F",
"M",
"M",
"F",
"M",
"M",
"F",
])
positive_label_index = pd.Series([
0, 1, 1, 0, 1, 0, 1, 1, 1, 1, 0, 1, 1, 1, 0, 0, 0, 0, 1, 0, 0, 1, 1, 0
])
group_variable = pd.Series([
1, 0, 2, 2, 1, 1, 2, 1, 1, 2, 0, 1, 2, 0, 1, 1, 1, 2, 0, 1, 0, 0, 1, 1
])
response = metric_one_vs_all(
CDDL,
x,
positive_label_index=positive_label_index == 1,
group_variable=group_variable)
assert response["F"] == approx(0.3982142857)
assert response["M"] == approx(-0.3982142857)