python基础2

'''

date：2018年6月21日

data：

purpose：

两张图，第一张横坐标轴是datetime 第二张图横坐标轴是省份

查看sku 销量 省份



注意点：

1、第一张图把省份做成list作区分

2、第二张图，做list的时候要先把string转换成float

3、转换方法 先把这列作为索引（也要保证唯一） 然后.to_dict()，然后指向这个索引位置即可





'''

import os

import sys

import pandas as pd

from datetime import date,datetime

import numpy as np

import matplotlib.pylab as plt

import warnings

warnings.filterwarnings('ignore')



dir_path = 'E:\\Data\\'

wdata = pd.read_csv(dir_path+'order_jx_y3',sep='\x01',encoding='utf-8',header=None)

wdata = wdata[[1,2,3,4,5]]

wdata.columns = ['province','city','sku','sku_qty','date']

wdata['date2'] = pd.to_datetime(wdata.date)

wdata['day'] = pd.to_datetime(wdata.date2).dt.dayofweek

mask = (wdata['date2'] > '2016-05-15') & (wdata['date2']< '2018-01-01')

mask1=(wdata['day']==1)

sku_info = pd.read_excel(dir_path+'/sku_info.xlsx',header=1,

                         names=['company','sku_no','item_name','length','width','height','vol','weight','class','level1','level2','level3','promotion'],

                         dtype={'company':str,'sku_no':str,'length':np.float64,'width':np.float64,'height':np.float64,'vol':np.float64,'weight':np.float64,'class':int,'promotion':int},

                         encoding='utf-8')



#不同省份不同sku的销量随时间变换

g1 = wdata.loc[mask & mask1]

plot_data = g1[['province','sku_qty','date2']].groupby(['province','date2']).sum()

plot_data = plot_data.reset_index()

plot_data = plot_data.sort_values(by=['province','date2'])

#plot_data.head(5)

province_list = plot_data['province'].drop_duplicates().tolist()



for i, province in enumerate(province_list):

    plt.style.use('ggplot')

    mask_plot = plot_data['province'] == province

    #plt.figure()

    plt.legend('province')

    plt.xlabel('date')

    plt.ylabel('sku_qty')

    plt.plot(plot_data['date2'].loc[mask_plot],plot_data['sku_qty'].loc[mask_plot])

#p = plot_data.groupby(['date2']).sum()

#plt.plot(p.index,p['sku_qty'])

#plt.savefig("E:\\Data\\allprovince.jpg")

plt.show()



#不同省份的sku销量

df=g1.copy()

tmp = df['sku'].drop_duplicates().reset_index().drop('index',axis=1).reset_index().set_index('sku')

dict_map = tmp['index'].to_dict()

df['sku'] = df['sku'].map(dict_map)

tmp1 = df['province'].drop_duplicates().reset_index().drop('index',axis=1).reset_index().set_index('province')

dict_map1 = tmp1['index'].to_dict()

df['province'] = df['province'].map(dict_map1)

plot_data1 = df[['province','sku_qty','sku']].groupby(['province','sku']).sum()

plot_data1 = plot_data1.reset_index()

plot_data1 = plot_data1.sort_values(by=['province','sku'])

sku_list = plot_data1['sku'].drop_duplicates().tolist()

plot_data1.head(2)

for i, sku in enumerate(sku_list):

    mask_plot1 = plot_data1['sku'] == sku

    plt.xlabel('province')

    plt.ylabel('sku_qty')

    plt.plot(plot_data1['province'].loc[mask_plot1],plot_data1['sku_qty'].loc[mask_plot1])

plt.show()