比较共享单车各用户类别的平均骑行时间趋势

学习目标：

掌握使用布尔型数组进行数据过滤

掌握如何构造ndarry

熟悉Matplotlib中图像元素的相关操作

一、明确任务

比较共享单车各用户类别（会员/非会员）的平均骑行时间

二、数据过滤

Numpy中广播操作，Numpy会将标量扩展成与要比较的数组同形状的一个数组，然后就可以一一对比。

一列数据与一个数据匹配：广播

 共享单车会员/非会员的平均骑行时间

import numpy as np
import matplotlib.pyplot as plt
import os

data_path = 'D:/BaiduNetdiskDownload/MOBIKE_CUP_2017/'
data_filenames =['test.csv','train.csv']

#数据收集+数据清洗
def collect_process_data():
    data_arr_list = []
    for data_filename in data_filenames:
        data_file = os.path.join(data_path,data_filename)
        data_arr = np.loadtxt(data_file,delimiter=',',dtype='str',skiprows=1)#读取数据
        #去掉双引号
        cln_data_arr = np.core.defchararray.replace(data_arr,'"','')
        data_arr_list.append(cln_data_arr)
    return data_arr_list
    
#数据分析
def get_mean_duration_by_type(data_arr_list,member_type):
    mean_duration = []
    #一列数据与一个数据匹配：广播
    for data_arr in data_arr_list:
        bool_arr = data_arr[:,-1] == member_type
        filtered_arr = data_arr[bool_arr]#过滤操作
        mean_duration = np.mean(filtered_arr[:,0].astype('float') / 1000 / 60 )
    return mean_duration

#结果展示
def save_show_result(Member_mean_list,Casual_mean_list):
    #1信息输出
    for idx in range(len(Member_mean_list)):
        Member_mean_duration = Member_mean_list[idx]
        Casual_mean_duration = Casual_mean_list[idx]
        print('第{}个季度，会员平均骑行时长：{:.2f}分钟，非会员平均骑行时长：{:.2f}分钟'.format(idx,Member_mean_duration,Casual_mean_duration))
    
    #2.构造多维数组  转置transpose()
    mean_duration_arr = np.array(Member_mean_duration,Casual_mean_duration).transpose()
    np.savetxt('./mean_duration.csv',mean_duration_arr,delimiter=',',header='Member,Casual',fmt='%.4f',comments='')
    
    #3.结果可视化
    plt.figure()
    plt.plot(Member_mean_duration, color='g',linestyle='-',maker ='o',lable ='Member')
    plt.plot(Casual_mean_duration, color='g',linestyle='--',maker ='*',lable ='Casual')
    plt.title('Member_vs_Casual')
    plt.xticks(range(0,4),[1,2,3,4],rotation=45)
    plt.tight_layout()
    plt.legend()
    plt.show()

#主函数
def main():
    #数据获取
    data_arr_list = collect_process_data()
    
    #数据分析

    Member_mean_list = get_mean_duration_by_type(data_arr_list,'Member')
    Casual_mean_list = get_mean_duration_by_type(data_arr_list,'Casual')
    #结果展示
    save_show_result(Member_mean_list,Casual_mean_list)

if __name__ == '__main__':
    main()