#레이블 인코딩
from sklearn.preprocessing import LabelEncoder
items = ['TV', '냉장고', '전자레인지', '컴퓨터', '선풍기', '선풍기', '믹서', '믹서']
#fit과 transform()으로 레이블 인코딩 수행
encoder = LabelEncoder()
encoder.fit(items)
labels = encoder.transform(items)
print('인코딩 변환값', labels)
#그러나 레이블인코딩은 ML알고리즘에서 가중치가 더 부여되거나 숫자에따라 더 중요하게 인식할 가능성이 큼
인코딩 변환값 [0 1 4 5 3 3 2 2]
#그래서 우리는 One-hot-Encoding을 사용하면 해결 가능하다.
from sklearn.preprocessing import OneHotEncoder
import numpy as np
items = ['TV', '냉장고', '전자레인지', '컴퓨터', '선풍기', '선풍기', '믹서', '믹서']
#먼저 숫자 값으로 변환을 위해 LabelEncoder로 변환
encoder = LabelEncoder()
encoder.fit(items)
labels = encoder.transform(items)
#2차원 데이터로 변환합니다.
labels = labels.reshape(-1,1)
#원핫 인코딩 해보장
oh_encoder = OneHotEncoder()
oh_encoder.fit(labels)
oh_labels = oh_encoder.transform(labels)
print('원핫 인코딩 데이터')
print(oh_labels.toarray())
print('원핫 인코딩 데이터 차원')
print(oh_labels.shape)
원핫 인코딩 데이터
[[1. 0. 0. 0. 0. 0.]
[0. 1. 0. 0. 0. 0.]
[0. 0. 0. 0. 1. 0.]
[0. 0. 0. 0. 0. 1.]
[0. 0. 0. 1. 0. 0.]
[0. 0. 0. 1. 0. 0.]
[0. 0. 1. 0. 0. 0.]
[0. 0. 1. 0. 0. 0.]]
원핫 인코딩 데이터 차원
(8, 6)
#판다스에는 원핫인코딩을 더 쉽게 지원하는 API가 있습니다.
#get_dummies()를 이용하면 됩니다.
import pandas as pd
df = pd.DataFrame({'item':['TV', '냉장고','전자레인지','컴퓨터', '선풍기','선풍기', '믹서', '믹서']})
print(df)
pd.get_dummies(df)
item
0 TV
1 냉장고
2 전자레인지
3 컴퓨터
4 선풍기
5 선풍기
6 믹서
7 믹서
item_TV item_냉장고 item_믹서 item_선풍기 item_전자레인지 item_컴퓨터
0 1 0 0 0 0 0
1 0 1 0 0 0 0
2 0 0 0 0 1 0
3 0 0 0 0 0 1
4 0 0 0 1 0 0
5 0 0 0 1 0 0
6 0 0 1 0 0 0
7 0 0 1 0 0 0
#피처스케일링과 정규화
#대표적인 방법으로 표준화와 정규화가 있다.
#표준화는 데이터 피처의 각각 평균이 0이고 분산이 1인 가우시안 정규 분포를 가진 값으로 변환하는 것을 의미한다.
#이 값은 원래 값에서 피처 x의 평균을 뺸 값을 x의 표준 편차로 나눈 값으로 계산 할 수 있다.
#정규화는
#다른 피처의 크기를 통일하기 위해 크기를 변환해주는 개념.
#최소 0~ 1까지의 값으로 변환하는 것을 의미
#새로우 데이터는 x피처의 최솟값을 뺀 값을 피처 x 최대값과 최소값의 차이로 나눈 값으로 변환 할 수 있습니다.
#그러나 사이런킷에서 제공하는 정규화는 약간 차이가 있다. Nomalizer 모듈은 선형대수에서의 정규화 개념이 적용됐으며 , 개별 백터의
#백터의 크기로 나눠준다. 혼선을 방지하기 위해 일반적인 의미의 표준화와 정규화 피처 스케일링으로 통칭
#선형대수 개념의 정규화를 백터 정규화로 지잋한다.
#대표적인 피처 스케일링 클래스인 StandardScaler와 MinMaxScaler를 알아보자
StandardScaler
standardScaler는 앞에서 설명한 표준화를 지원하기 위한 클래스이다.
RBF커널을 이용하는 서포트 백터 머신, 선형 회귀, 로지스틱 회귀는
데이터가 가우시안 분포를 가지고 있다고 구현 됐기에
사전에 표준화를 적용하는 것은 예측 성능 향상에 중요한 요소가 될 수 있습니다.
from sklearn.datasets import load_iris
import pandas as pd
iris = load_iris()
iris_data = iris.data
iris_df = pd.DataFrame(data=iris_data, columns=iris.feature_names)
print(iris_df.head())
print('feature 들의 평균 값')
print(iris_df.mean())
print('\n feature들의 분산 값')
print(iris_df.var())
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm)
0 5.1 3.5 1.4 0.2
1 4.9 3.0 1.4 0.2
2 4.7 3.2 1.3 0.2
3 4.6 3.1 1.5 0.2
4 5.0 3.6 1.4 0.2
feature 들의 평균 값
sepal length (cm) 5.843333
sepal width (cm) 3.057333
petal length (cm) 3.758000
petal width (cm) 1.199333
dtype: float64
feature들의 분산 값
sepal length (cm) 0.685694
sepal width (cm) 0.189979
petal length (cm) 3.116278
petal width (cm) 0.581006
dtype: float64
StandardScaler를 이용해 각 피처들을 한 번에 표준화해 변화해보자
fit()과 transform() 메서드에 변환 대상 피처 데이터 세트를 입력하고 호출하면 간단하게 변환된다.
transform() 스케일 변환된 데이터 세트가 넘파이의 ndarray이므로 이를 DataFrame으로 변환해 평균값과 분산 값을 확인
from sklearn.preprocessing import StandardScaler
# StandardScaler 객체 생성
scaler = StandardScaler()
# StandardScaler로 데이터 세트 변환 fit transform 호출
scaler.fit(iris_df)
iris_scaled = scaler.transform(iris_df)
iris_scaled
array([[-9.00681170e-01, 1.01900435e+00, -1.34022653e+00,
-1.31544430e+00],
[-1.14301691e+00, -1.31979479e-01, -1.34022653e+00,
-1.31544430e+00],
[-1.38535265e+00, 3.28414053e-01, -1.39706395e+00,
-1.31544430e+00],
[-1.50652052e+00, 9.82172869e-02, -1.28338910e+00,
-1.31544430e+00],
[-1.02184904e+00, 1.24920112e+00, -1.34022653e+00,
-1.31544430e+00],
[-5.37177559e-01, 1.93979142e+00, -1.16971425e+00,
-1.05217993e+00],
[-1.50652052e+00, 7.88807586e-01, -1.34022653e+00,
-1.18381211e+00],
[-1.02184904e+00, 7.88807586e-01, -1.28338910e+00,
-1.31544430e+00],
[-1.74885626e+00, -3.62176246e-01, -1.34022653e+00,
-1.31544430e+00],
[-1.14301691e+00, 9.82172869e-02, -1.28338910e+00,
-1.44707648e+00],
[-5.37177559e-01, 1.47939788e+00, -1.28338910e+00,
-1.31544430e+00],
[-1.26418478e+00, 7.88807586e-01, -1.22655167e+00,
-1.31544430e+00],
[-1.26418478e+00, -1.31979479e-01, -1.34022653e+00,
-1.44707648e+00],
[-1.87002413e+00, -1.31979479e-01, -1.51073881e+00,
-1.44707648e+00],
[-5.25060772e-02, 2.16998818e+00, -1.45390138e+00,
-1.31544430e+00],
[-1.73673948e-01, 3.09077525e+00, -1.28338910e+00,
-1.05217993e+00],
[-5.37177559e-01, 1.93979142e+00, -1.39706395e+00,
-1.05217993e+00],
[-9.00681170e-01, 1.01900435e+00, -1.34022653e+00,
-1.18381211e+00],
[-1.73673948e-01, 1.70959465e+00, -1.16971425e+00,
-1.18381211e+00],
[-9.00681170e-01, 1.70959465e+00, -1.28338910e+00,
-1.18381211e+00],
[-5.37177559e-01, 7.88807586e-01, -1.16971425e+00,
-1.31544430e+00],
[-9.00681170e-01, 1.47939788e+00, -1.28338910e+00,
-1.05217993e+00],
[-1.50652052e+00, 1.24920112e+00, -1.56757623e+00,
-1.31544430e+00],
[-9.00681170e-01, 5.58610819e-01, -1.16971425e+00,
-9.20547742e-01],
[-1.26418478e+00, 7.88807586e-01, -1.05603939e+00,
-1.31544430e+00],
[-1.02184904e+00, -1.31979479e-01, -1.22655167e+00,
-1.31544430e+00],
[-1.02184904e+00, 7.88807586e-01, -1.22655167e+00,
-1.05217993e+00],
[-7.79513300e-01, 1.01900435e+00, -1.28338910e+00,
-1.31544430e+00],
[-7.79513300e-01, 7.88807586e-01, -1.34022653e+00,
-1.31544430e+00],
[-1.38535265e+00, 3.28414053e-01, -1.22655167e+00,
-1.31544430e+00],
[-1.26418478e+00, 9.82172869e-02, -1.22655167e+00,
-1.31544430e+00],
[-5.37177559e-01, 7.88807586e-01, -1.28338910e+00,
-1.05217993e+00],
[-7.79513300e-01, 2.40018495e+00, -1.28338910e+00,
-1.44707648e+00],
[-4.16009689e-01, 2.63038172e+00, -1.34022653e+00,
-1.31544430e+00],
[-1.14301691e+00, 9.82172869e-02, -1.28338910e+00,
-1.31544430e+00],
[-1.02184904e+00, 3.28414053e-01, -1.45390138e+00,
-1.31544430e+00],
[-4.16009689e-01, 1.01900435e+00, -1.39706395e+00,
-1.31544430e+00],
[-1.14301691e+00, 1.24920112e+00, -1.34022653e+00,
-1.44707648e+00],
[-1.74885626e+00, -1.31979479e-01, -1.39706395e+00,
-1.31544430e+00],
[-9.00681170e-01, 7.88807586e-01, -1.28338910e+00,
-1.31544430e+00],
[-1.02184904e+00, 1.01900435e+00, -1.39706395e+00,
-1.18381211e+00],
[-1.62768839e+00, -1.74335684e+00, -1.39706395e+00,
-1.18381211e+00],
[-1.74885626e+00, 3.28414053e-01, -1.39706395e+00,
-1.31544430e+00],
[-1.02184904e+00, 1.01900435e+00, -1.22655167e+00,
-7.88915558e-01],
[-9.00681170e-01, 1.70959465e+00, -1.05603939e+00,
-1.05217993e+00],
[-1.26418478e+00, -1.31979479e-01, -1.34022653e+00,
-1.18381211e+00],
[-9.00681170e-01, 1.70959465e+00, -1.22655167e+00,
-1.31544430e+00],
[-1.50652052e+00, 3.28414053e-01, -1.34022653e+00,
-1.31544430e+00],
[-6.58345429e-01, 1.47939788e+00, -1.28338910e+00,
-1.31544430e+00],
[-1.02184904e+00, 5.58610819e-01, -1.34022653e+00,
-1.31544430e+00],
[ 1.40150837e+00, 3.28414053e-01, 5.35408562e-01,
2.64141916e-01],
[ 6.74501145e-01, 3.28414053e-01, 4.21733708e-01,
3.95774101e-01],
[ 1.28034050e+00, 9.82172869e-02, 6.49083415e-01,
3.95774101e-01],
[-4.16009689e-01, -1.74335684e+00, 1.37546573e-01,
1.32509732e-01],
[ 7.95669016e-01, -5.92373012e-01, 4.78571135e-01,
3.95774101e-01],
[-1.73673948e-01, -5.92373012e-01, 4.21733708e-01,
1.32509732e-01],
[ 5.53333275e-01, 5.58610819e-01, 5.35408562e-01,
5.27406285e-01],
[-1.14301691e+00, -1.51316008e+00, -2.60315415e-01,
-2.62386821e-01],
[ 9.16836886e-01, -3.62176246e-01, 4.78571135e-01,
1.32509732e-01],
[-7.79513300e-01, -8.22569778e-01, 8.07091462e-02,
2.64141916e-01],
[-1.02184904e+00, -2.43394714e+00, -1.46640561e-01,
-2.62386821e-01],
[ 6.86617933e-02, -1.31979479e-01, 2.51221427e-01,
3.95774101e-01],
[ 1.89829664e-01, -1.97355361e+00, 1.37546573e-01,
-2.62386821e-01],
[ 3.10997534e-01, -3.62176246e-01, 5.35408562e-01,
2.64141916e-01],
[-2.94841818e-01, -3.62176246e-01, -8.98031345e-02,
1.32509732e-01],
[ 1.03800476e+00, 9.82172869e-02, 3.64896281e-01,
2.64141916e-01],
[-2.94841818e-01, -1.31979479e-01, 4.21733708e-01,
3.95774101e-01],
[-5.25060772e-02, -8.22569778e-01, 1.94384000e-01,
-2.62386821e-01],
[ 4.32165405e-01, -1.97355361e+00, 4.21733708e-01,
3.95774101e-01],
[-2.94841818e-01, -1.28296331e+00, 8.07091462e-02,
-1.30754636e-01],
[ 6.86617933e-02, 3.28414053e-01, 5.92245988e-01,
7.90670654e-01],
[ 3.10997534e-01, -5.92373012e-01, 1.37546573e-01,
1.32509732e-01],
[ 5.53333275e-01, -1.28296331e+00, 6.49083415e-01,
3.95774101e-01],
[ 3.10997534e-01, -5.92373012e-01, 5.35408562e-01,
8.77547895e-04],
[ 6.74501145e-01, -3.62176246e-01, 3.08058854e-01,
1.32509732e-01],
[ 9.16836886e-01, -1.31979479e-01, 3.64896281e-01,
2.64141916e-01],
[ 1.15917263e+00, -5.92373012e-01, 5.92245988e-01,
2.64141916e-01],
[ 1.03800476e+00, -1.31979479e-01, 7.05920842e-01,
6.59038469e-01],
[ 1.89829664e-01, -3.62176246e-01, 4.21733708e-01,
3.95774101e-01],
[-1.73673948e-01, -1.05276654e+00, -1.46640561e-01,
-2.62386821e-01],
[-4.16009689e-01, -1.51316008e+00, 2.38717193e-02,
-1.30754636e-01],
[-4.16009689e-01, -1.51316008e+00, -3.29657076e-02,
-2.62386821e-01],
[-5.25060772e-02, -8.22569778e-01, 8.07091462e-02,
8.77547895e-04],
[ 1.89829664e-01, -8.22569778e-01, 7.62758269e-01,
5.27406285e-01],
[-5.37177559e-01, -1.31979479e-01, 4.21733708e-01,
3.95774101e-01],
[ 1.89829664e-01, 7.88807586e-01, 4.21733708e-01,
5.27406285e-01],
[ 1.03800476e+00, 9.82172869e-02, 5.35408562e-01,
3.95774101e-01],
[ 5.53333275e-01, -1.74335684e+00, 3.64896281e-01,
1.32509732e-01],
[-2.94841818e-01, -1.31979479e-01, 1.94384000e-01,
1.32509732e-01],
[-4.16009689e-01, -1.28296331e+00, 1.37546573e-01,
1.32509732e-01],
[-4.16009689e-01, -1.05276654e+00, 3.64896281e-01,
8.77547895e-04],
[ 3.10997534e-01, -1.31979479e-01, 4.78571135e-01,
2.64141916e-01],
[-5.25060772e-02, -1.05276654e+00, 1.37546573e-01,
8.77547895e-04],
[-1.02184904e+00, -1.74335684e+00, -2.60315415e-01,
-2.62386821e-01],
[-2.94841818e-01, -8.22569778e-01, 2.51221427e-01,
1.32509732e-01],
[-1.73673948e-01, -1.31979479e-01, 2.51221427e-01,
8.77547895e-04],
[-1.73673948e-01, -3.62176246e-01, 2.51221427e-01,
1.32509732e-01],
[ 4.32165405e-01, -3.62176246e-01, 3.08058854e-01,
1.32509732e-01],
[-9.00681170e-01, -1.28296331e+00, -4.30827696e-01,
-1.30754636e-01],
[-1.73673948e-01, -5.92373012e-01, 1.94384000e-01,
1.32509732e-01],
[ 5.53333275e-01, 5.58610819e-01, 1.27429511e+00,
1.71209594e+00],
[-5.25060772e-02, -8.22569778e-01, 7.62758269e-01,
9.22302838e-01],
[ 1.52267624e+00, -1.31979479e-01, 1.21745768e+00,
1.18556721e+00],
[ 5.53333275e-01, -3.62176246e-01, 1.04694540e+00,
7.90670654e-01],
[ 7.95669016e-01, -1.31979479e-01, 1.16062026e+00,
1.31719939e+00],
[ 2.12851559e+00, -1.31979479e-01, 1.61531967e+00,
1.18556721e+00],
[-1.14301691e+00, -1.28296331e+00, 4.21733708e-01,
6.59038469e-01],
[ 1.76501198e+00, -3.62176246e-01, 1.44480739e+00,
7.90670654e-01],
[ 1.03800476e+00, -1.28296331e+00, 1.16062026e+00,
7.90670654e-01],
[ 1.64384411e+00, 1.24920112e+00, 1.33113254e+00,
1.71209594e+00],
[ 7.95669016e-01, 3.28414053e-01, 7.62758269e-01,
1.05393502e+00],
[ 6.74501145e-01, -8.22569778e-01, 8.76433123e-01,
9.22302838e-01],
[ 1.15917263e+00, -1.31979479e-01, 9.90107977e-01,
1.18556721e+00],
[-1.73673948e-01, -1.28296331e+00, 7.05920842e-01,
1.05393502e+00],
[-5.25060772e-02, -5.92373012e-01, 7.62758269e-01,
1.58046376e+00],
[ 6.74501145e-01, 3.28414053e-01, 8.76433123e-01,
1.44883158e+00],
[ 7.95669016e-01, -1.31979479e-01, 9.90107977e-01,
7.90670654e-01],
[ 2.24968346e+00, 1.70959465e+00, 1.67215710e+00,
1.31719939e+00],
[ 2.24968346e+00, -1.05276654e+00, 1.78583195e+00,
1.44883158e+00],
[ 1.89829664e-01, -1.97355361e+00, 7.05920842e-01,
3.95774101e-01],
[ 1.28034050e+00, 3.28414053e-01, 1.10378283e+00,
1.44883158e+00],
[-2.94841818e-01, -5.92373012e-01, 6.49083415e-01,
1.05393502e+00],
[ 2.24968346e+00, -5.92373012e-01, 1.67215710e+00,
1.05393502e+00],
[ 5.53333275e-01, -8.22569778e-01, 6.49083415e-01,
7.90670654e-01],
[ 1.03800476e+00, 5.58610819e-01, 1.10378283e+00,
1.18556721e+00],
[ 1.64384411e+00, 3.28414053e-01, 1.27429511e+00,
7.90670654e-01],
[ 4.32165405e-01, -5.92373012e-01, 5.92245988e-01,
7.90670654e-01],
[ 3.10997534e-01, -1.31979479e-01, 6.49083415e-01,
7.90670654e-01],
[ 6.74501145e-01, -5.92373012e-01, 1.04694540e+00,
1.18556721e+00],
[ 1.64384411e+00, -1.31979479e-01, 1.16062026e+00,
5.27406285e-01],
[ 1.88617985e+00, -5.92373012e-01, 1.33113254e+00,
9.22302838e-01],
[ 2.49201920e+00, 1.70959465e+00, 1.50164482e+00,
1.05393502e+00],
[ 6.74501145e-01, -5.92373012e-01, 1.04694540e+00,
1.31719939e+00],
[ 5.53333275e-01, -5.92373012e-01, 7.62758269e-01,
3.95774101e-01],
[ 3.10997534e-01, -1.05276654e+00, 1.04694540e+00,
2.64141916e-01],
[ 2.24968346e+00, -1.31979479e-01, 1.33113254e+00,
1.44883158e+00],
[ 5.53333275e-01, 7.88807586e-01, 1.04694540e+00,
1.58046376e+00],
[ 6.74501145e-01, 9.82172869e-02, 9.90107977e-01,
7.90670654e-01],
[ 1.89829664e-01, -1.31979479e-01, 5.92245988e-01,
7.90670654e-01],
[ 1.28034050e+00, 9.82172869e-02, 9.33270550e-01,
1.18556721e+00],
[ 1.03800476e+00, 9.82172869e-02, 1.04694540e+00,
1.58046376e+00],
[ 1.28034050e+00, 9.82172869e-02, 7.62758269e-01,
1.44883158e+00],
[-5.25060772e-02, -8.22569778e-01, 7.62758269e-01,
9.22302838e-01],
[ 1.15917263e+00, 3.28414053e-01, 1.21745768e+00,
1.44883158e+00],
[ 1.03800476e+00, 5.58610819e-01, 1.10378283e+00,
1.71209594e+00],
[ 1.03800476e+00, -1.31979479e-01, 8.19595696e-01,
1.44883158e+00],
[ 5.53333275e-01, -1.28296331e+00, 7.05920842e-01,
9.22302838e-01],
[ 7.95669016e-01, -1.31979479e-01, 8.19595696e-01,
1.05393502e+00],
[ 4.32165405e-01, 7.88807586e-01, 9.33270550e-01,
1.44883158e+00],
[ 6.86617933e-02, -1.31979479e-01, 7.62758269e-01,
7.90670654e-01]])
transfrom() 시 스케일 변환된 데이터 세트가 NumPy ndarray로 변환돼 이를 DataFrame으로 변환
iris_df_scaled = pd.DataFrame(data=iris_scaled, columns=iris.feature_names)
iris_df_scaled
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm)
0 -0.900681 1.019004 -1.340227 -1.315444
1 -1.143017 -0.131979 -1.340227 -1.315444
2 -1.385353 0.328414 -1.397064 -1.315444
3 -1.506521 0.098217 -1.283389 -1.315444
4 -1.021849 1.249201 -1.340227 -1.315444
... ... ... ... ...
145 1.038005 -0.131979 0.819596 1.448832
146 0.553333 -1.282963 0.705921 0.922303
147 0.795669 -0.131979 0.819596 1.053935
148 0.432165 0.788808 0.933271 1.448832
149 0.068662 -0.131979 0.762758 0.790671
150 rows × 4 columns
#feature들의 평균값
iris_df_scaled.mean()
sepal length (cm) -1.690315e-15
sepal width (cm) -1.842970e-15
petal length (cm) -1.698641e-15
petal width (cm) -1.409243e-15
dtype: float64
#feature들의 분산값
iris_df_scaled.var()
sepal length (cm) 1.006711
sepal width (cm) 1.006711
petal length (cm) 1.006711
petal width (cm) 1.006711
dtype: float64
MinMaxScaler
가우시안 분포가 아닐 경우에 적용해볼 수 있다
from sklearn.preprocessing import MinMaxScaler
#minMaxScaler객체 생성
scaler = MinMaxScaler()
# MinMaxSclaer로 데이터 세트 변환 fit() transform() 호출
scaler.fit(iris_df)
iris_scaled = scaler.transform(iris_df)
iris_df_scaled
sepal length (cm) sepal width (cm) petal length (cm) petal width (cm)
0 -0.900681 1.019004 -1.340227 -1.315444
1 -1.143017 -0.131979 -1.340227 -1.315444
2 -1.385353 0.328414 -1.397064 -1.315444
3 -1.506521 0.098217 -1.283389 -1.315444
4 -1.021849 1.249201 -1.340227 -1.315444
... ... ... ... ...
145 1.038005 -0.131979 0.819596 1.448832
146 0.553333 -1.282963 0.705921 0.922303
147 0.795669 -0.131979 0.819596 1.053935
148 0.432165 0.788808 0.933271 1.448832
149 0.068662 -0.131979 0.762758 0.790671
150 rows × 4 columns
#feature들의 최소값
iris_df_scaled.min()
sepal length (cm) -1.870024
sepal width (cm) -2.433947
petal length (cm) -1.567576
petal width (cm) -1.447076
dtype: float64
#feature들의 최대값
iris_df_scaled.max()
sepal length (cm) 2.492019
sepal width (cm) 3.090775
petal length (cm) 1.785832
petal width (cm) 1.712096
dtype: float64
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
%matplotlib inline
titanic_df = pd.read_csv('titanic_train.csv')
titanic_df.head(3)
PassengerId Survived Pclass Name Sex Age SibSp Parch Ticket Fare Cabin Embarked
0 1 0 3 Braund, Mr. Owen Harris male 22.0 1 0 A/5 21171 7.2500 NaN S
1 2 1 1 Cumings, Mrs. John Bradley (Florence Briggs Th... female 38.0 1 0 PC 17599 71.2833 C85 C
2 3 1 3 Heikkinen, Miss. Laina female 26.0 0 0 STON/O2. 3101282 7.9250 NaN S
# 학습데이터 정보
print(titanic_df.info())
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 891 entries, 0 to 890
Data columns (total 12 columns):
# Column Non-Null Count Dtype
--- ------ -------------- -----
0 PassengerId 891 non-null int64
1 Survived 891 non-null int64
2 Pclass 891 non-null int64
3 Name 891 non-null object
4 Sex 891 non-null object
5 Age 714 non-null float64
6 SibSp 891 non-null int64
7 Parch 891 non-null int64
8 Ticket 891 non-null object
9 Fare 891 non-null float64
10 Cabin 204 non-null object
11 Embarked 889 non-null object
dtypes: float64(2), int64(5), object(5)
memory usage: 83.7+ KB
None
titanic_df['Age'].fillna(titanic_df['Age'].mean(), inplace=True)
titanic_df['Cabin'].fillna('N', inplace=True)
titanic_df['Embarked'].fillna('N', inplace=True)
print('데이터세트 NULL값 개수', titanic_df.isnull().sum().sum())
데이터세트 NULL값 개수 0
print('sex값의 분포 :\n', titanic_df['Sex'].value_counts())
print('\n Canbin 값 분포 : \n', titanic_df['Cabin'].value_counts())
print('\n Embarked 값 분포: \n', titanic_df['Embarked'].value_counts())
sex값의 분포 :
male 577
female 314
Name: Sex, dtype: int64
Canbin 값 분포 :
N 687
C23 C25 C27 4
G6 4
B96 B98 4
F2 3
...
C32 1
D49 1
A32 1
D15 1
A36 1
Name: Cabin, Length: 148, dtype: int64
Embarked 값 분포:
S 644
C 168
Q 77
N 2
Name: Embarked, dtype: int64
titanic_df['Cabin'] = titanic_df['Cabin'].str[:1]
print(titanic_df['Cabin'].head(3))
0 N
1 C
2 N
Name: Cabin, dtype: object
titanic_df.groupby(['Sex', 'Survived'])['Survived'].count()
Sex Survived
female 0 81
1 233
male 0 468
1 109
Name: Survived, dtype: int64
sns.barplot(x='Sex', y = 'Survived', data = titanic_df)
<AxesSubplot:xlabel='Sex', ylabel='Survived'>
sns.barplot(x='Pclass', y='Survived', hue='Sex', data=titanic_df)
#입력 age에 따라 구분 값을 반환하는 함수 설정 , DataFrame의 apply lamda식에 사용
def get_category(age):
cat = ''
if age <= -1: cat = 'UnKnown'
elif age <= 5 : cat = 'Baby'
elif age <= 12 : cat = 'Child'
elif age <= 18 : cat = 'Teenager'
elif age <= 25 : cat = 'Student'
elif age <= 35 : cat = 'Young Adult'
elif age <= 60 : cat = 'Adult'
else : cat = 'Elderly'
return cat
# 막대그래프의 크기 figure를 더 크게 설정
plt.figure(figsize=(10,6))
#X축의 값을 순차적으로 표시하기 위한 설정
group_names = ['Unknown', 'Baby', 'Child', 'Teenager', 'Student', 'Young Adult', 'Adult', 'Elderly']
#lambda식 위에서 생성한 get_category() 함수를 반환값으로 지정.
#get_category(X)는 입력값으로 'Age'칼럼 값을 받아서 해당하는 cat 반환
titanic_df['Age_cat'] = titanic_df['Age'].apply(lambda x : get_category(x))
sns.barplot(x='Age_cat', y = 'Survived', hue = 'Sex', data= titanic_df, order=group_names)
titanic_df.drop('Age_cat', axis =1, inplace = True)
from sklearn import preprocessing
def encode_features(dataDF):
features = ['Cabin', 'Sex', 'Embarked']
for feature in features:
le = preprocessing.LabelEncoder()
le = le.fit(dataDF[feature])
dataDF[feature] = le.transform(dataDF[feature])
return dataDF
titanic_df = encode_features(titanic_df)
titanic_df.head()
