#!/usr/bin/env python3

# -*- coding: utf-8 -*-

"""

Created on Fri Feb  3 09:37:21 2023

@author: ming

"""

import pandas as pd

from fancyimpute import KNN

from sklearn.neighbors import KNeighborsClassifier

from sklearn.naive_bayes import GaussianNB

from sklearn.linear_model import LogisticRegression

from sklearn.svm import SVC

from sklearn.tree import DecisionTreeClassifier

from sklearn.ensemble import RandomForestClassifier

from sklearn.neural_network import MLPClassifier

from sklearn.model_selection import train_test_split

from sklearn.metrics import classification_report,confusion_matrix

from sklearn import preprocessing

from sklearn.metrics import roc_auc_score

from sklearn.model_selection import GridSearchCV

#load data

X=pd.read_csv(r"/Users/ming/Downloads/Xset1.csv")

Y=pd.read_csv(r"/Users/ming/Downloads/Y.csv")

#del columns

X.dropna(thresh=100)

#impute

Xc=X.columns

X = pd.DataFrame(KNN(k=6).fit_transform(X)) 

X.columns=Xc

#standardize

standardized_X = preprocessing.scale(X)

#oversampling

#imblearn进行随机过采样

from imblearn.over_sampling import RandomOverSampler

ros = RandomOverSampler(random_state=0)

X_resampled, y_resampled = ros.fit_resample(standardized_X, Y)

models = {

    "knn": KNeighborsClassifier(n_neighbors=1),

    "naive_bayes": GaussianNB(),

    "logit": LogisticRegression(solver="lbfgs", multi_class="auto"),

    "svm": SVC(kernel="rbf", gamma="auto"),

    "decision_tree": DecisionTreeClassifier(criterion='gini',splitter='random',max_depth=15,min_samples_leaf=10,min_samples_split=10), 

    "random_forest": RandomForestClassifier(n_estimators=60,max_depth=6,max_samples=70,max_features=5),

    "mlp": MLPClassifier()

}

trainX, testX, trainY, testY = train_test_split(X_resampled, y_resampled, random_state=3, test_size=0.2)

print("use '{}' bulid model...".format("random_forest"))

model = models["random_forest"]

model.fit(trainX, trainY)

'''

tree_param_grid = {'n_estimators': range(50,100,10),

                  'min_samples_split':range(30,111,20),

                  'max_features':range(5,76,10),

                  'max_depth':range(3,22,3)

                  }

tree_best = GridSearchCV(model,param_grid = tree_param_grid,cv = 3,scoring="roc_auc",n_jobs= -1, verbose = 100) 

#Gridsearch turning parameters  

tree_best.fit(trainX, trainY)

print(tree_best.best_estimator_)

print(tree_best.best_params_,"  ","得分：",tree_best.best_score_)

'''

features_importance = pd.concat([pd.DataFrame(Xc).T,pd.DataFrame(model.feature_importances_).T],axis=0).T

features_importance.columns=["feature","importance"]

## prediction

print("elvaluation...")

predictions = model.predict(testX)

print(classification_report(testY, predictions))

auc_score = roc_auc_score(testY, predictions)

print('CONFUSION MATRIX')

print(confusion_matrix(testY, predictions))

print("AUC")

print(auc_score)