def test_default_transform():
bpsketch = BinningProcessSketch(variable_names)
bpsketch.add(df, y)
with raises(NotFittedError):
bpsketch.transform(df, metric="woe")
bpsketch.solve()
with raises(TypeError):
X_transform = bpsketch.transform(df.values, metric="woe")
with raises(ValueError):
X_transform = bpsketch.transform(df, metric="new_woe")
X_transform = bpsketch.transform(df)
optb = OptimalBinningSketch()
x = df["mean radius"]
optb.add(x, y)
optb.solve()
assert optb.transform(x,
metric="woe") == approx(X_transform["mean radius"],
rel=1e-6)
def test_numerical_default_tdigest():
optb = OptimalBinningSketch(sketch="t-digest", eps=1e-4)
optb.add(x, y)
optb.solve()
assert optb.status == "OPTIMAL"
optb.binning_table.build()
assert optb.binning_table.iv == approx(5.04392547, rel=1e-2)
optb.binning_table.analysis()
assert optb.binning_table.gini == approx(0.87541620, rel=1e-2)
assert optb.binning_table.js == approx(0.39378376, rel=1e-2)
assert optb.binning_table.quality_score == approx(0.0, rel=1e-2)
def test_categorical_default_user_splits():
x = np.array([
'Working', 'State servant', 'Working', 'Working', 'Working',
'State servant', 'Commercial associate', 'State servant', 'Pensioner',
'Working', 'Working', 'Pensioner', 'Working', 'Working', 'Working',
'Working', 'Working', 'Working', 'Working', 'State servant', 'Working',
'Commercial associate', 'Working', 'Pensioner', 'Working', 'Working',
'Working', 'Working', 'State servant', 'Working',
'Commercial associate', 'Working', 'Working', 'Commercial associate',
'State servant', 'Working', 'Commercial associate', 'Working',
'Pensioner', 'Working', 'Commercial associate', 'Working', 'Working',
'Pensioner', 'Working', 'Working', 'Pensioner', 'Working',
'State servant', 'Working', 'State servant', 'Commercial associate',
'Working', 'Commercial associate', 'Pensioner', 'Working', 'Pensioner',
'Working', 'Working', 'Working', 'Commercial associate', 'Working',
'Pensioner', 'Working', 'Commercial associate', 'Commercial associate',
'State servant', 'Working', 'Commercial associate',
'Commercial associate', 'Commercial associate', 'Working', 'Working',
'Working', 'Commercial associate', 'Working', 'Commercial associate',
'Working', 'Working', 'Pensioner', 'Working', 'Pensioner', 'Working',
'Working', 'Pensioner', 'Working', 'State servant', 'Working',
'Working', 'Working', 'Working', 'Working', 'Commercial associate',
'Commercial associate', 'Commercial associate', 'Working',
'Commercial associate', 'Working', 'Working', 'Pensioner'
],
dtype=object)
y = np.array([
1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 1, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0,
0, 0, 0, 0
])
optb = OptimalBinningSketch(dtype="categorical",
solver="mip",
cat_cutoff=0.1,
verbose=True)
optb.add(x, y)
optb.solve()
assert optb.status == "OPTIMAL"
def test_verbose():
optb = OptimalBinningSketch(verbose=True)
optb.add(x, y)
optb.solve()
assert optb.status == "OPTIMAL"
def test_information():
optb = OptimalBinningSketch(solver="cp")
with raises(NotFittedError):
optb.information()
optb.add(x, y)
optb.solve()
with raises(ValueError):
optb.information(print_level=-1)
optb.information(print_level=0)
optb.information(print_level=1)
optb.information(print_level=2)
optb = OptimalBinningSketch(solver="mip")
optb.add(x, y)
optb.solve()
optb.information(print_level=2)
def test_params():
with raises(TypeError):
OptimalBinningSketch(name=1)
with raises(ValueError):
OptimalBinningSketch(dtype="nominal")
with raises(ValueError):
OptimalBinningSketch(sketch="new_sketch")
with raises(ValueError):
OptimalBinningSketch(eps=-1e-2)
with raises(ValueError):
OptimalBinningSketch(K=-3)
with raises(ValueError):
OptimalBinningSketch(solver="new_solver")
with raises(ValueError):
OptimalBinningSketch(divergence="new_divergence")
with raises(ValueError):
OptimalBinningSketch(max_n_prebins=-2)
with raises(ValueError):
OptimalBinningSketch(min_n_bins=-2)
with raises(ValueError):
OptimalBinningSketch(max_n_bins=-2.2)
with raises(ValueError):
OptimalBinningSketch(min_n_bins=3, max_n_bins=2)
with raises(ValueError):
OptimalBinningSketch(min_bin_size=0.6)
with raises(ValueError):
OptimalBinningSketch(max_bin_size=-0.6)
with raises(ValueError):
OptimalBinningSketch(min_bin_size=0.5, max_bin_size=0.3)
with raises(ValueError):
OptimalBinningSketch(min_bin_n_nonevent=-2)
with raises(ValueError):
OptimalBinningSketch(max_bin_n_nonevent=-2)
with raises(ValueError):
OptimalBinningSketch(min_bin_n_nonevent=3, max_bin_n_nonevent=2)
with raises(ValueError):
OptimalBinningSketch(min_bin_n_event=-2)
with raises(ValueError):
OptimalBinningSketch(max_bin_n_event=-2)
with raises(ValueError):
OptimalBinningSketch(min_bin_n_event=3, max_bin_n_event=2)
with raises(ValueError):
OptimalBinningSketch(monotonic_trend="new_trend")
with raises(ValueError):
OptimalBinningSketch(min_event_rate_diff=1.1)
with raises(ValueError):
OptimalBinningSketch(max_pvalue=1.1)
with raises(ValueError):
OptimalBinningSketch(max_pvalue_policy="new_policy")
with raises(ValueError):
OptimalBinningSketch(gamma=-0.2)
with raises(ValueError):
OptimalBinningSketch(cat_cutoff=-0.2)
with raises(TypeError):
OptimalBinningSketch(cat_heuristic=1)
with raises(TypeError):
OptimalBinningSketch(special_codes={1, 2, 3})
with raises(ValueError):
OptimalBinningSketch(split_digits=9)
with raises(ValueError):
OptimalBinningSketch(mip_solver="new_solver")
with raises(ValueError):
OptimalBinningSketch(time_limit=-2)
with raises(TypeError):
OptimalBinningSketch(verbose=1)
def test_numerical_default_tdigest_merge():
optb1 = OptimalBinningSketch(sketch="t-digest", eps=1e-4)
optb2 = OptimalBinningSketch(sketch="t-digest", eps=1e-4)
x1, x2 = x[:200], x[200:]
y1, y2 = y[:200], y[200:]
optb1.add(x1, y1)
optb2.add(x2, y2)
optb1.merge(optb2)
optb1.solve()
assert optb1.status == "OPTIMAL"
optb1.binning_table.build()
assert optb1.binning_table.iv == approx(5.04392547, rel=1e-2)
optb1.binning_table.analysis()
assert optb1.binning_table.gini == approx(0.87541620, rel=1e-2)
assert optb1.binning_table.js == approx(0.39378376, rel=1e-2)
assert optb1.binning_table.quality_score == approx(0.0, rel=1e-2)