将dataset按比例随机划分训练集和测试集

通常一个数据集合并没有划分为training set 和 test set，而为了减少过拟合，就需要我们自己对数据集进行划分

索性写了一个python函数方便任何比例的划分，其中每个样本的选取是随机的（不重复选）

"""
divide the original data set into training set and test set
percent  -- percentage of training set

"""

import numpy as np
import random as rd


def divideTrainAndTest():
    # python 2.x needs raw_input()
    filename = input("Enter the file name:")
    o_data = np.loadtxt(filename)
    o_rows = o_data.shape[0]
    o_columns = o_data.shape[1]
    # require input of trainingset percent
    percent = float(input("Enter the percentage of training data:(0~1)"))
    tr_rows = int(percent * o_rows)
    ts_rows = o_rows - tr_rows
    tr_data = np.zeros([tr_rows, o_columns], dtype=list)
    ts_data = np.zeros([ts_rows, o_columns], dtype=list)
    # get a bool array to record whether a sample is in the training set
    i = 0
    isInTrain = []
    while i < o_rows:
        isInTrain.append(False)
        i = i+1
    # get the training set
    i = 0
    while i < tr_rows:
        j = rd.randint(0,o_rows-1)
        if isInTrain[j] == False:
            isInTrain[j] = True
            k = 0
            while k < o_columns:
                tr_data[i][k] = o_data[j][k]
                k = k+1
            i = i+1
    # get the test set
    i = 0
    j = 0
    while i