双十一美妆

import numpy as np
import pandas as pd
df=pd.read_csv('D:/data/双十一淘宝美妆数据.csv')
df

df.info()#查看数值类型，日期类型

df['update_time'].unique()#以数组形式返回唯一值

df.columns

df.describe()

df.isnull().sum()

df.fillna(0,inplace=True)#空值填充

df.duplicated().sum()

df=df.drop_duplicates()
df.reset_index(drop=True,inplace=True)
#重建索引，drop=True;把原来的索引index列去掉，丢掉，（重新排序）
df

df.duplicated().sum()

## 日期格式化，格式为“%Y-%m-%d” 

#df['update_time'] = pd.to_datetime(df['update_time']).apply(lambda x:x.strftime('%Y-%m-%d'))
df['update_time'] = pd.to_datetime(df['update_time'])
df['update_time'] = df['update_time'].apply(lambda x:x.strftime('%Y-%m-%d'))
df.info()

df

## 增加一列销售额

df['sale_amount'] = df['price']*df['sale_count']
df[df['sale_count']>0].sort_values(by='sale_count')
#检查一下最小值，最小值最好计算sale_amount是否准确


result1=df.groupby('update_time')['sale_count'].sum().to_dict()
#list(result1.keys())
result2=list(round(float(x/10e6),2) for x in result1.values())

import pyecharts.options as opts
from pyecharts.charts import Line
dayXL=(
    Line()
    .add_xaxis(xaxis_data=list(result1.keys()))
    .add_yaxis(
        y_axis=result2,
        series_name="销售量(单位：千万件)"
    )
    .set_global_opts(
        yaxis_opts=opts.AxisOpts(
            splitline_opts=opts.SplitLineOpts(
                is_show=True,
            ),
            name="销量",
        ),
        xaxis_opts=opts.AxisOpts(
            name="时间",
        ),
    )
)
dayXL.render('html/dayXL.html')

## 按照日期统计出累计销售量排行TOP10的店铺 

datetimes = list(df['update_time'].unique())#unique()以数组的形式返回唯一值，并按从小到大的顺序排序
datetimes.reverse()
datetimes

temp1 = df[df['update_time']<=datetimes[0]].groupby('店名').agg({'sale_count':'sum','sale_amount':'sum'})
temp2 = temp1.sort_values(by='sale_count',ascending=False)[:10].sort_values(by='sale_count')
#temp2.to_dict()['sale_count'].values()
label = list(temp2.to_dict()['sale_count'].keys())
saleCount = list(temp2.to_dict()['sale_count'].values())
saleAmount = list(temp2.to_dict()['sale_amount'].values())

#result1=df.groupby('update_time')['sale_count'].sum().to_dict()
#list(result1.keys())
#result2=list(round(float(x/10e6),2) for x in result1.values())
#

from pyecharts import options as opts
from pyecharts.charts import Bar, Timeline

timeline=(
    Timeline()
    .add_schema(
        is_loop_play=True,
        play_interval=500
    )
)
for datetime in datetimes:
    temp1 = df[df['update_time']<=datetime].groupby('店名').agg({'sale_count':'sum','sale_amount':'sum'})
    temp2 = temp1.sort_values(by='sale_count',ascending=False)[:10].sort_values(by='sale_count')
    temp3 = temp1.sort_values(by='sale_count',ascending=False)[:10].sort_values(by='sale_amount')
    label = list(temp2.to_dict()['sale_count'].keys())
    
    
    temp4 = temp2.to_dict()['sale_count']
    saleCount = list(round(float(x/10e5),2) for x in temp4.values())
    #saleCount = list(temp2.to_dict()['sale_count'].values())
    temp5 = temp2.to_dict()['sale_amount']
    saleAmount = list(round(float(x/10e6),2) for x in temp5.values())
    #saleAmount = list(temp2.to_dict()['sale_amount'].values())
    bar=(
        Bar()
        .add_xaxis(label)
        .add_yaxis(series_name="销售量(单位：百万件)",y_axis=saleCount,label_opts=opts.LabelOpts(position="right"))
        .add_yaxis(series_name="销售额(单位：千万)",y_axis=saleAmount,label_opts=opts.LabelOpts(position="right"))
        .set_global_opts(
            title_opts=opts.TitleOpts(title="累计销售量排行TOP10的店铺")
        )
        .reversal_axis()#x轴与y轴转置,变成条形图
    )
    timeline.add(bar,datetime)
timeline.render('html/timelineChart.html')

from pyecharts import options as opts
from pyecharts.charts import Bar, Timeline

timeline=(
    Timeline()
    .add_schema(
        is_loop_play=True,
        play_interval=500
    )
)
for datetime in datetimes:
    temp1 = df[df['update_time']<=datetime].groupby('店名').agg({'price':'mean','sale_amount':'mean'})
    
    
    temp2 = temp1.sort_values(by='price',ascending=False)[:20].sort_values(by='price')
    #temp3 = temp1.sort_values(by='price',ascending=False)[:20].sort_values(by='sale_amount')
    label = list(temp2.to_dict()['price'].keys())
    salePrice = list( round(float(x),2) for x in temp2.to_dict()['price'].values())
    
    bar=(
        Bar()
        .add_xaxis(label)
       # .add_yaxis(series_name="销售量(单位：百万件)",y_axis=saleCount,label_opts=opts.LabelOpts(position="right"))
        .add_yaxis(series_name="平均价格(单位：元/件)",y_axis=salePrice,label_opts=opts.LabelOpts(position="right"))
        .set_global_opts(
            title_opts=opts.TitleOpts(title="平均价格排行TOP20的店铺")
        )
        .reversal_axis()#x轴与y轴转置,变成条形图
    )
    timeline.add(bar,datetime)
timeline.render('html/Price.html')