泰坦尼克号旅客生存预测记录

import urllib.request
import os

url = "xxxxx"

data_path = "D:/xxx"

if not os.path.isfile(data_path):　　# 如果不存在文件
    ret = urllib.request.urlretrieve(url, data_path)　　#则下载
    print("Download: ", ret)
else:
    print(data_path, "already exists")

df = pd.read_excel(data_path) #读取文件
select_cols = ["survived", "name", "pcalss", "sex", "age", "sibsp", "parch", "fare", "embarked"] #选出要用的列
df = df[select_cols]
df.isnull()#可列出所有存在空或者NA的行，为一个True和False的矩阵，若元素为空或者NA则为True，否则为False



df.isnull().any()

df.isnull.sum()

df[df.isnull().values == True]

填充缺失值，用平均值填充

avg_age = df[‘age‘].mean()

df[‘age‘] = df[‘age‘].fillna(avg_age)

将性别转换成序号

df[‘sex‘] = df[‘sex‘].map({‘female‘ : 0, ‘male‘ : 1}).astype(int)

删除一列

axis = 1表示删除列

df = df.drop([‘name‘], axis = 1)

数据标准化

from sklearn import preprocessing

minmax_scale = preprocessing.MinMaxScaler(frature_range = (0, 1))

x_data = minmax_scale.fit_transform(x_data)