import pandas as pd

import matplotlib.pyplot as plt

import seaborn as sns

 

train = pd.read_csv('train.csv')

test = pd.read_csv('test.csv')

 

train = train.drop(['intensity'], axis =1)

 

def feature_change(train):

    # type

    train['type'] = train['type'].map({'white': 0, 'red' : 1})

    

    #sweetness

    train['sweetness'] = train['sweetness'].map({'dry': 0, 'off-dry' : 1, 'medium-sweet' : 2})

    return train

 

train = feature_change(train)

 

test = test.drop(['intensity'], axis =1)

 

def feature_change(test):

    # type

    test['type'] = test['type'].map({'white': 0, 'red' : 1})

    

    #sweetness

    test['sweetness'] = test['sweetness'].map({'dry': 0, 'off-dry' : 1, 'medium-sweet' : 2})

    return test

test = feature_change(test)

 

train.info()

plt.figure(figsize = (12,12))

sns.heatmap(data = train.corr(), annot =True)

 

 

이상치 제거는 있다가 해보겠음