import pandas as pd

import seaborn as sns

import numpy as np

import matplotlib.pyplot as plt 

 

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

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

 

train.columns

test.columns

 

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

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

 

train.isnull().sum()

 

# intensity 제거 

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)

 

# 이상치만 잘 다듬는다면... 

# 이 대회가 새로운 모델을 만들라는건 아니잖아? 

# 보는 눈(해석)이 중요하다 

train = train[train['density'] < 1.00128]

train = train[train['chlorides']< 0.3]

 

train.info()

train.shape

 

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

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

train_x = train.drop(columns=['quality'])

train_y = train['quality']

 

 

from sklearn.model_selection import train_test_split

train_x, val_x, train_y, val_y  = train_test_split(train_x, train_y, test_size=0.2, random_state=42)

 

print(train_x.shape)

print(val_x.shape)

print(train_y.shape)

print(val_y.shape)

 

from sklearn.ensemble import RandomForestClassifier

from sklearn.metrics import confusion_matrix

from sklearn.metrics import classification_report

 

model = RandomForestClassifier(n_estimators=375, max_depth=14, random_state=42)

model.fit(train.drop(columns='quality'),train['quality'])

 

y_pred = model.predict(val_x)

print(confusion_matrix(val_y, y_pred))

print(classification_report(val_y, y_pred))

 

submission = pd.read_csv('sample_submission.csv')

y_pred = model.predict(test)

submission['quality'] = y_pred

 

submission.to_csv('submission_labtop5.csv', index=False)

점검 필요