toad 模型结果保存、读取及评分

        在使用toad的时候没有看到这个model如何保存模型和读取模型，最后在根据结果进行评分的模块，所以就写一写。

栗子：

加载所需model

import pandas as pd
import numpy as np
import toad as td
import scorecardpy as sc
from sklearn.datasets import load_iris

导入iris数据并处理成dataframe格式：

df = pd.DataFrame(iris.data, columns=iris.feature_names)

# 添加目标列
df['target'] = iris.target

df['target_names'] = iris.target_names[df['target']]

#去除不必要feature
feature_list = list(df)

feature_list.remove('target_names')


#筛选有用特征
train_selected, dropped = td.selection.select(df[feature_list],target = 'target', empty = 0.5, iv = 0.05, corr = 0.9, return_drop=True)

woe转化：

# 初始化
transer = td.transform.WOETransformer()


# combiner.transform() & transer.fit_transform() 转化训练数据，并去掉target列
train_woe = transer.fit_transform(c.transform(train_selected), train_selected['target'], exclude=['target'])


# 将woe转化后的数据做逐步回归
final_data = td.selection.stepwise(train_woe,target = 'target', estimator='ols', direction = 'both', criterion = 'aic')

col = list(final_data.drop(['target'],axis=1).columns)

模型训练：

card = td.ScoreCard(
    combiner = c,
    transer = transer,
    #class_weight = 'balanced',
    #C=0.1,
    #base_score = 600,
    #base_odds = 35 ,
    #pdo = 60,
    #rate = 2
)

card.fit(final_data[col], final_data['target'])

结果保存、读取：

#结果保存
np.save('iris_test',card.export(to_frame=True))

#结果读取
card_load = np.load('iris_test.npy',allow_pickle=True)


#对columns进行修改 因为要用scorecard的model来读取toad的评分卡
scard = pd.DataFrame(card_load,columns=['variable','bin','points'])

#修饰一下toad评分卡的格式
scard['bin'] = scard['bin'].apply(lambda x: x.replace('~',','))
scard['bin'] = scard['bin'].apply(lambda x: x.replace(' ',''))

构建数据进行评分

#构建数据
aa = [[3.3,3.3,3.3]]

test_data = pd.DataFrame(aa,columns=['sepal width (cm)','sepal length (cm)','petal width (cm)'])

#评分
sc.scorecard_ply(test_data,scard)

评分卡以及评分结果