# -*- coding: utf-8 -*-
# @ author hulei 2016-5-3
from numpy import *
import operator
from os import listdir
 
import sys
reload(sys)
sys.setdefaultencoding('utf8')
 
# x,y=getDataSet_dz('iris.data.txt',4)
 
def getDataSet(filename,numberOfFeature):  #將數(shù)據(jù)集讀入內(nèi)存 
 fr = open(filename)
 numberOfLines = len(fr.readlines())   #get the number of lines in the file file.readlines()是把文件的全部內(nèi)容讀到內(nèi)存，并解析成一個list
 returnMat = zeros((numberOfLines,numberOfFeature))  #prepare matrix to return 3代表數(shù)據(jù)集中特征數(shù)目###
 classLabelVector = []      #prepare labels return 
 fr = open(filename)
 index = 0
 for line in fr.readlines():
  line = line.strip()     #strip() 參數(shù)為空時，默認(rèn)刪除空白符（包括'\n', '\r', '\t', ' ')
  listFromLine = line.split(',')   #split 以什么為標(biāo)準(zhǔn)分割一次 分成數(shù)組中的每個元素
  returnMat[index,:] = listFromLine[0:numberOfFeature] 
  #classLabelVector.append(int(listFromLine[-1])) #append() 方法向列表的尾部添加一個新的元素
  if listFromLine[-1] == 'Iris-setosa' :
   classLabelVector.append(1)
  elif listFromLine[-1] == 'Iris-versicolor' :
   classLabelVector.append(2)
  else:
  #elif listFromLine[-1] == 'Iris-virginica' :
   classLabelVector.append(3)
  index += 1
 return returnMat,classLabelVector
 
def getDataSet_dz(filename,numberOfFeature): #改進版，可以消除數(shù)據(jù)中的空白行
 numberOfLines = 0
 mx = []  #將數(shù)據(jù)集 去除空行后存入
 fr = open(filename)
 for line in fr.readlines():  
  line = line.strip() 
  if line != '' : #去除空白行 
   numberOfLines+=1
   mx.append( line.split(',') )
 returnMat = zeros((numberOfLines,numberOfFeature))
 classLabelVector = [] 
 for index in range(numberOfLines) :
  returnMat[index,:] = mx[index][0:numberOfFeature] 
  if mx[index][-1] == 'Iris-setosa' :
   classLabelVector.append(1)
  elif mx[index][-1] == 'Iris-versicolor' :
   classLabelVector.append(2)
  else:
  #elif listFromLine[-1] == 'Iris-virginica' :
   classLabelVector.append(3)
 return returnMat,classLabelVector