import numpy as np
import pandas as pd
import os
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LogisticRegression
from sklearn.model_selection import RepeatedKFold
from sklearn.model_selection import GridSearchCV
import matplotlib.pyplot as plt

path = os.getcwd()
os.chdir('/Users/marin/TEACHING/2324/M2-GLM-HAX912X/TP')

data = pd.read_csv("voice.csv")
data.info()

x = data.drop(["label"], axis = 1)
x.info()


scaler = StandardScaler().fit(x)
x = scaler.transform(x)
y = data.label

x_train, x_test, y_train, y_test = train_test_split(x, y, 
  test_size = 0.07, random_state = 42)
x_test.shape

logit = LogisticRegression(penalty=None)
logit.fit(x_train,y_train)
y_chap = logit.predict(x_test)
table=pd.crosstab(y_test,y_chap)
table

Crossvalid = RepeatedKFold(n_splits=5, n_repeats=5)
parameters = {'C':np.linspace(0.1,2.1,num=11)}

logitl1 = GridSearchCV(LogisticRegression(penalty='l1',solver='saga',
                                          max_iter=1000), parameters,
                       cv=Crossvalid,scoring='neg_log_loss')
logitl1.fit(x_train,y_train)
logitl1.best_params_
logitl1.best_score_
results = pd.DataFrame(logitl1.cv_results_)
results.mean_test_score

plt.plot(np.linspace(0.1,2.1,num=11),results.mean_test_score)
plt.show()

y_chap = logitl1.predict(x_test)
table=pd.crosstab(y_test,y_chap)
table

logitl2 = GridSearchCV(LogisticRegression(penalty='l2',solver='lbfgs',
                                          max_iter=1000), parameters,
                       cv=Crossvalid,scoring='neg_log_loss')
logitl2.fit(x_train,y_train)
logitl2.best_params_
logitl2.best_score_
results = pd.DataFrame(logitl2.cv_results_)
results.mean_test_score

plt.plot(np.linspace(0.1,2.1,num=11),results.mean_test_score)
plt.show()

y_chap = logitl2.predict(x_test)
table=pd.crosstab(y_test,y_chap)
table