python - 如何使用 Pandas 中的变量之一绘制堆积条形图？

Question

我正在尝试使用这个 csv 文件，该文件已作为 pandas.Dataframe 输入，为各种购物者提供黑色星期五购买数据以及用于了解他们的购买模式的各种变量。

User_ID,Product_ID,Gender,Age,Occupation,City_Category,Stay_In_Current_City_Years,Marital_Status,Product_Category_1,Product_Category_2,Product_Category_3,Purchase
1000001,P00069042,F,0-17,10,A,2,0,3,,,8370
1000001,P00248942,F,0-17,10,A,2,0,1,6,14,15200
1000001,P00087842,F,0-17,10,A,2,0,12,,,1422
1000001,P00085442,F,0-17,10,A,2,0,12,14,,1057
1000002,P00285442,M,55+,16,C,4+,0,8,,,7969
1000003,P00193542,M,26-35,15,A,3,0,1,2,,15227
1000004,P00184942,M,46-50,7,B,2,1,1,8,17,19215
1000004,P00346142,M,46-50,7,B,2,1,1,15,,15854
1000004,P0097242,M,46-50,7,B,2,1,1,16,,15686
1000005,P00274942,M,26-35,20,A,1,1,8,,,7871
1000005,P00251242,M,26-35,20,A,1,1,5,11,,5254
1000005,P00014542,M,26-35,20,A,1,1,8,,,3957
1000005,P00031342,M,26-35,20,A,1,1,8,,,6073
1000005,P00145042,M,26-35,20,A,1,1,1,2,5,15665
1000006,P00231342,F,51-55,9,A,1,0,5,8,14,5378
1000006,P00190242,F,51-55,9,A,1,0,4,5,,2079
1000006,P0096642,F,51-55,9,A,1,0,2,3,4,13055
1000006,P00058442,F,51-55,9,A,1,0,5,14,,8851
1000007,P00036842,M,36-45,1,B,1,1,1,14,16,11788
1000008,P00249542,M,26-35,12,C,4+,1,1,5,15,19614
1000008,P00220442,M,26-35,12,C,4+,1,5,14,,8584
1000008,P00156442,M,26-35,12,C,4+,1,8,,,9872
1000008,P00213742,M,26-35,12,C,4+,1,8,,,9743
1000008,P00214442,M,26-35,12,C,4+,1,8,,,5982
1000008,P00303442,M,26-35,12,C,4+,1,1,8,14,11927
1000009,P00135742,M,26-35,17,C,0,0,6,8,,16662
1000009,P00039942,M,26-35,17,C,0,0,8,,,5887
1000009,P00161442,M,26-35,17,C,0,0,5,14,,6973
1000009,P00078742,M,26-35,17,C,0,0,5,8,14,5391
1000010,P00085942,F,36-45,1,B,4+,1,2,4,8,16352
1000010,P00118742,F,36-45,1,B,4+,1,5,11,,8886
1000010,P00297942,F,36-45,1,B,4+,1,8,,,5875
1000010,P00266842,F,36-45,1,B,4+,1,5,,,8854
1000010,P00058342,F,36-45,1,B,4+,1,3,4,,10946
1000010,P00032442,F,36-45,1,B,4+,1,5,,,5152
1000010,P00105942,F,36-45,1,B,4+,1,5,,,7089
1000010,P00182642,F,36-45,1,B,4+,1,2,4,9,12909
1000010,P00186942,F,36-45,1,B,4+,1,5,12,,8770
1000010,P00155442,F,36-45,1,B,4+,1,1,11,15,15212
1000010,P00221342,F,36-45,1,B,4+,1,1,2,5,15705
1000010,P00087242,F,36-45,1,B,4+,1,14,,,7947
1000010,P00111142,F,36-45,1,B,4+,1,1,15,16,18963
1000010,P00259342,F,36-45,1,B,4+,1,5,9,,8718
1000010,P0094542,F,36-45,1,B,4+,1,2,4,9,16406
1000010,P00148642,F,36-45,1,B,4+,1,6,10,13,12642
1000010,P00312142,F,36-45,1,B,4+,1,8,,,10007
1000010,P00113242,F,36-45,1,B,4+,1,1,6,8,11562

现在我想创建一个按城市和性别划分的总购买量的堆积图，如下所示:这是我尝试过的:

import pandas
import matplotlib.pyplot as plt
from matplotlib.ticker import StrMethodFormatter
import numpy as np
with open('BlackFriday.csv') as csv_file:
    df = pandas.read_csv(csv_file, sep=',')
    # Group by user id, city and gender
    users_by_city_gender = df.groupby(['City_Category','Gender'])['Purchase'].agg('sum').to_frame()
    ax3 = pandas.DataFrame({'City-A': users_by_city_gender.groupby('City_Category').get_group('A').Purchase,
                            'City-B': users_by_city_gender.groupby('City_Category').get_group('B').Purchase,
                            'City-C': users_by_city_gender.groupby('City_Category').get_group('C').Purchase}).plot.hist(stacked=True)
    ## Switch off ticks
    ax3.tick_params(axis="both", which="both", bottom=False, top=False, labelbottom=False, left=False, right=False,
                    labelleft=True)

    # Draw horizontal axis lines
    # vals = ax.get_yticks()
    # for tick in vals:
    #    ax.axhline(y=tick, linestyle='dashed', alpha=0.4, color='#eeeeee', zorder=1)

    # Remove title
    ax3.set_title("Total purchase by city and gender")

    # Set x-axis label
    ax3.set_xlabel("City category", labelpad=20, weight='bold', size=12)

    # Set y-axis label
    ax3.set_ylabel("Total purchase [dollars]", labelpad=20, weight='bold', size=12)

    # Format y-axis label
    ax3.yaxis.set_major_formatter(StrMethodFormatter('{x:,g}'))
    plt.show()

结果图是 这似乎与我想要的情节完全不同。调试 users_by_city_gender 显示它是一系列城市(A、B 和 C)的数据框，每个城市包含按性别(M 和 F)划分的总购买量。所以我认为这就是正确绘制图表所需的数据。

我已经查看了 stackexchange 上有关为 pandas 数据框创建堆积条形图的其他问题，但我无法找到解决我的问题的方法。

Answer 1

您可以使用groupby和pivot_table:

s = (df.pivot_table(
        index='City_Category', columns='Gender', values='Purchase', aggfunc='sum'))

s.plot(kind='bar', stacked=True)
plt.show()

<小时/>

为了便于说明，pivot 的结果如下所示:

Gender                F         M
City_Category
A               55412.0   54047.0
B              201995.0   62543.0
C                   NaN  108604.0