[note] e-commerce order data analysis practice

One 、 E-commerce order data analysis practice

1.1 Data cleaning

# %load prep.py
import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

plt.rcParams['font.sans-serif'] = ['SimHei']
plt.rcParams['axes.unicode_minus'] = False

%config InlineBackend.figure_format = 'svg'

ec_df = pd.read_csv('datas/ecomm/ An e-commerce platform 2021 Annual order data .csv', index_col=0)

#  transformation orderTime and payTime by datetime
ec_df['orderTime'] = pd.to_datetime(ec_df['orderTime'])
ec_df['payTime'] = pd.to_datetime(ec_df['payTime'])
ec_df

#  Extract orderTime stay 2021 Years of data 
ec_df = ec_df[ec_df['orderTime'].dt.year == 2021]
ec_df

#  Extract payTime No later than orderTime 30 Minute data 、 Payment amount and order amount are greater than 0 The data of 
ec_df = ec_df[(ec_df['payTime'] - ec_df['orderTime']).dt.total_seconds() <= 1800]
ec_df = ec_df[ec_df['payTime'] >= ec_df['orderTime']]
ec_df = ec_df[(ec_df['payment'] >= 0) & (ec_df['orderAmount'] >= 0)]
ec_df

#  fill chanelID Column null value 
ec_df['chanelID'] = ec_df['chanelID'].fillna(ec_df['chanelID'].mode()[0])
ec_df.info()

#  Rename the channel column and platform column 
ec_df = ec_df.rename(columns={
    'chanelID': 'channelID', 'platfromType': 'platformType'})
ec_df

#  Change the typo of the platform type 
ser = ec_df['platformType'].replace(r'[\s·]', '', regex=True)
ser = ser.str.title()
ec_df['platformType'] = ser.replace([' Weixin ', 'Vx', ' Webpage ', ' Website '], 
                                    [' WeChat ', ' WeChat ', 'Web', 'Web'])

#  take payment Greater than orderAmount Change the data of 
temp_df = ec_df[ec_df['payment'] <= ec_df['orderAmount']]
mean_discount = (temp_df['payment'] / temp_df['orderAmount']).mean()
ec_df['payment'] = np.round(
    ec_df['payment'].where(
    	ec_df['payment'] <= ec_df['orderAmount'],
    	ec_df['orderAmount'] * mean_discount
	)
)

1.2 Data analysis

1.2.1 Calculate the overall indicators

#  Calculation GMV
GMV = ec_df['orderAmount'].sum()
print(GMV)

#  Calculate total sales 
all_payment = ec_df['payment'].sum()
print(all_payment)

#  Calculate actual sales 
true_payment = ec_df[ec_df['chargeback'] == False]['payment'].sum()
print(true_payment)

#  Calculate return rate 
chargeback_ratio = f"{
      100 * len(ec_df[ec_df['chargeback']]) / len(ec_df):.2f}%"
print(chargeback_ratio)

#  Calculate the customer unit price 
ARPU = true_payment / ec_df['userID'].nunique()
print(ARPU)

1.2.2 Calculate monthly GMV And trend analysis

#  Add month column 
ec_df['month'] = ec_df['orderTime'].dt.month
ec_df

#  Add the actual payment amount column 
temp_df = ec_df
temp_df['true_payment'] = temp_df['payment'].where(ec_df['chargeback'] == False, 0)

#  Generate month pairs GMV And actual payment pivot table 
monthly_gmv = pd.pivot_table(temp_df, index='month', values=['orderAmount', 'true_payment'], aggfunc='sum')
monthly_gmv = monthly_gmv.applymap(lambda x: np.round(x/10000, 2))

#  Start drawing 
import pyecharts.options as opts
from pyecharts.charts import Line

gmv_line = Line()
#  Add month data 
gmv_line.add_xaxis(monthly_gmv.index.to_list())
#  Join in GMV data 
gmv_line.add_yaxis('GMV', monthly_gmv['orderAmount'].to_list())
#  Add net sales data 
gmv_line.add_yaxis(' Actual sales ', monthly_gmv['true_payment'].to_list())
#  Modify global configuration 
gmv_line.set_global_opts(
    title_opts=opts.TitleOpts(title="2021 Annual monthly sales figures "),
    yaxis_opts=opts.AxisOpts(name='GMV( Company ： Ten thousand yuan )'),
    xaxis_opts=opts.AxisOpts(name=' Time ( Company ： month )')
)

gmv_line.render_notebook()

1.2.3 Disassembly of flow channel sources GMV Proportion

#  Build platform 、GMV PivotTable 
channel_gmv = pd.pivot_table(temp_df, index='channelID', values=['orderAmount'], aggfunc='sum')
channel_gmv = channel_gmv.applymap(lambda x: x/ 10000)
channel_gmv

#  Start drawing 
from pyecharts.charts import Pie

#  Create pie chart objects 
channel_pie = Pie()
#  Add data , Set the size of the inner and outer rings of the pie chart 
channel_pie.add(
    '',
    channel_gmv.reset_index().values.tolist(),
    radius=["50%", "75%"]
)

#  Configure global settings 
channel_pie.set_global_opts(
        title_opts=opts.TitleOpts(title=" Various channels GMV Proportion "),
        legend_opts=opts.LegendOpts(orient="vertical", pos_top="15%", pos_left="2%")
)

#  Configure the percentage of values displayed on the label 
channel_pie.set_series_opts(label_opts=opts.LabelOpts(formatter="{b}: {d}%"))

channel_pie.render_notebook()

1.2.4 Count the orders by the day of the week

#  obtain orderTime The day of the week 
temp_df['weekday'] = temp_df['orderTime'].dt.weekday.map(lambda x: f' week {
      x + 1}')
temp_df

#  Generate a PivotTable of the day of the week and the order quantity 
weekday_count = pd.pivot_table(temp_df, index='weekday', values='orderID', aggfunc='nunique')
weekday_count

#  Start drawing 
from pyecharts.charts import Bar
from pyecharts.globals import ThemeType

#  Create a histogram object 
weekday_bar = Bar({
    "theme": ThemeType.LIGHT})
#  add to x Axis data 
weekday_bar.add_xaxis(weekday_count.index.tolist())
#  add to y Axis data 
weekday_bar.add_yaxis(" Order quantity ", weekday_count['orderID'].tolist())

#  Configure global settings with title 
weekday_bar.set_global_opts(title_opts=opts.TitleOpts(title=" Order quantity per week "))

#  Add marker lines 
weekday_bar.set_series_opts(
    label_opts=opts.LabelOpts(is_show=True),
    markline_opts=opts.MarkLineOpts(
        data=[
            opts.MarkLineItem(type_="min", name=" minimum value "),
            opts.MarkLineItem(type_="max", name=" Maximum ")
        ]
    ),
)

weekday_bar.render_notebook()

1.2.5 Count the order quantity according to the order period （30 Every minute ）

#  Create a time period column 
temp_df['hour'] = temp_df['orderTime'].dt.floor('30T').dt.strftime('%H:%M')
temp_df

#  Generate pivot table of time period and order quantity 
time_count = pd.pivot_table(temp_df, index='hour', values='orderID', aggfunc='nunique')
time_count

#  Create a histogram object 
weekday_bar = Bar({
    "theme": ThemeType.ESSOS})
#  add to x Axis data 
weekday_bar.add_xaxis(time_count.index.tolist())
#  add to y Axis data 
weekday_bar.add_yaxis(" Order quantity ", time_count['orderID'].tolist())

#  Configure global settings with title 
weekday_bar.set_global_opts(
    title_opts=opts.TitleOpts(title=" Order quantity in each period "),
    datazoom_opts=opts.DataZoomOpts()
)

#  Add marker lines 
weekday_bar.set_series_opts(
    label_opts=opts.LabelOpts(is_show=True),
    markline_opts=opts.MarkLineOpts(
        data=[
            opts.MarkLineItem(type_="min", name=" minimum value "),
            opts.MarkLineItem(type_="max", name=" Maximum ")
        ]
    ),
)

weekday_bar.render_notebook()

1.2.6 Calculate the repurchase rate by month

#  Generate each... By month userID Statistics are not repeated orderID My PivotTable 
multiple_bought = pd.pivot_table(temp_df, index='userID', columns='month', values='orderID', aggfunc='nunique')
multiple_bought = multiple_bought.applymap(
    lambda x: np.nan if np.isnan(x) else (
        0 if x == 1 else 1
    )
)
np.round(100 * multiple_bought.sum() / multiple_bought.count(), 2)

#  No more pictures , Laugh to death