def forward_iv_test():
df = pd.read_csv('credit_old.csv')
df = df[['Age', 'target']]
df = df.dropna()
t = forwardSplit(df['Age'], df['target'])
t.fit(sby='iv', minv=0.1, init_split=20)
print(
t.bins
) # [16. 25. 29. 33. 36. 38. 40. 42. 44. 46. 48. 50. 58. 60. 63. 94.]
t = forwardSplit(df['Age'], df['target'])
t.fit(sby='iv', num_split=4, init_split=20, min_sample=len(df) * 0.2)
print(t.bins) # [16. 38. 50. 94.]
t.fit(sby='woeiv', num_split=4, init_split=20)
print(t.bins) # [16. 25. 33. 36. 38. 94.]
print("bin\twoe")
for i in range(len(t.bins) - 1):
v = t.value[(t.x < t.bins[i + 1]) & (t.x >= t.bins[i])]
woe = t._cal_woe(v)
print((t.bins[i], t.bins[i + 1]), woe)
def forward_woe_test_cat():
df = pd.read_csv('credit_old.csv')
df = df[['Branch', 'target']]
df = df.dropna()
t = forwardSplit(df['Branch'], df['target'], categorical=True)
t.fit(sby='woe', minv=0.01, init_split=0, num_split=4)
print(t.bins)
for i in range(len(t.bins)):
v = t.value[np.isin(t.x_idx, t.bins[i])]
woe = t._cal_woe(v)
print(t.bins[i], woe)
def forward_iv_test2():
df = pd.read_csv('resolution.csv')
df = df.dropna()
t = forwardSplit(df['x'], df['y'], missing=-1)
t.fit(sby='woeiv', minv=0.1, init_split=0, num_split=4)
print(t.bins)
print("bin\twoe")
for i in range(len(t.bins) - 1):
v = t.value[(t.x < t.bins[i + 1]) & (t.x >= t.bins[i])]
woe = t._cal_woe(v)
print((t.bins[i], t.bins[i + 1]), woe)