import urllib.request
import requests
from io import StringIO
import numpy as np
import pandas as pd
'''
下載網(wǎng)絡文件，并導入CSV文件作為numpy的矩陣
'''
# 網(wǎng)絡數(shù)據(jù)文件地址
url = "http://archive.ics.uci.edu/ml/machine-learning-databases/pima-indians-diabetes/pima-indians-diabetes.data"
# 方法一
# ========================================================
# 下載文件
#r = urllib.request.urlopen(url)
# 導入CSV文件作為numpy的矩陣
#dataset = np.loadtxt(r, delimiter=",")
# 方法二
# ========================================================
# 下載文件
#r = requests.get(url)
# 導入CSV文件作為numpy的矩陣
#dataset = np.loadtxt(StringIO(r.text), delimiter=",") # 此處用到 StringIO ！?。。。?！
# 方法三
# ========================================================
#用genfromtxt直接下載網(wǎng)絡文件，并將CSV文件導作numpy矩陣。爽！?。。。。。?！
#dataset = np.genfromtxt(url, delimiter=",")
# 方法四
# ========================================================
# 用pandas.read_csv直接下載網(wǎng)絡文件，并將CSV文件導作pandas.DataFrame。
# dataset = pd.read_csv('http://www-bcf.usc.edu/~gareth/ISL/Advertising.csv', index_col=0)
dataset = pd.read_csv(url)
# ========================================================
# separate the data from the target attributes
X = dataset[:,0:7]
y = dataset[:,8]
print(X)
#print(y)