backward elimination.py

# -*- coding: utf-8 -*-
"""
Created on Thu Jan  2 15:56:30 2025

@author: asus
"""

import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
##true false gibi leri labelencoder iler daha fazla prametresi olani one hotencoder iler numeric hale getiryoruz
veriler = pd.read_csv('verihavadurumu.csv')
veriler
from sklearn import preprocessing 
veriler2=veriler.apply(preprocessing.LabelEncoder().fit_transform)
veriler2
c=veriler2.iloc[:,:1]
from sklearn import preprocessing
ohe=preprocessing.OneHotEncoder()
c=ohe.fit_transform(c).toarray()
c
havadurumu=pd.DataFrame(data=c,index=range(14),columns=['o','r','s'])
sonveriler=pd.concat([havadurumu,veriler.iloc[:,1:3]],axis=1)
sonveriler=pd.concat([veriler2.iloc[:,-2:],sonveriler],axis=1)
sonveriler
from sklearn.model_selection import train_test_split

x_train, x_test,y_train,y_test = train_test_split(sonveriler.iloc[:,:1],sonveriler.iloc[:,-1:],test_size=0.33, random_state=0)

from sklearn.linear_model import LinearRegression
regressor= LinearRegression()
regressor.fit(x_train,y_train)

y_pred = regressor.predict(x_test)

import statsmodels.api as sm
X=np.append(arr=np.ones((14,1)).astype(int),values=sonveriler.iloc[:,:-1],axis=1)
X_1=sonveriler.iloc[:,[0,1,2,3,4,5]].values
X_1=np.array(X_1,dtype=float)
model=sm.OLS(sonveriler.iloc[:,-1:],X_1).fit()
print(model.summary())

sonveriler=sonveriler.iloc[:,1:]



import statsmodels.api as sm
X=np.append(arr=np.ones((14,1)).astype(int),values=sonveriler.iloc[:,:-1],axis=1)

X_1=sonveriler.iloc[:,[0,1,2,3,4]].values
X_1=np.array(X_1,dtype=float)
model=sm.OLS(sonveriler.iloc[:,-1:],X_1).fit()
print(model.summary())


x_train=x_train.iloc[:,1:]
x_test=x_test.iloc[:,1:]
regressor.fit(x_train,y_train)