如何在python中绘制和显示数据集的分布？

Question

我有一个数据集，其中的一小部分如下所示，

data = [ ['2018-01-01',  1.323 ,    'AI' ,   2000,'Communications','Mothers'], 
   ['2018-01-02',  1.525 ,    'AI',    1500,'Communications','Mothers'],
   ['2018-01-03',  1.045 ,    'AI' ,    500,'Communications','Mothers'],
   ['2018-01-04',  1.845 ,    'AI' ,    600,'Communications','Mothers'],
  ['2018-01-05',  1.045 ,    'AI' ,    500,'Communications','Mothers'],
   ['2018-01-02',  1.446  ,  'BOC' ,    550,'Pharmaceuticals','JASDAQ Standard'],
   ['2018-01-03',  2.110 ,   'BOC' ,   3201,'Pharmaceuticals','JASDAQ Standard'],
   ['2018-01-04',  2.150 ,   'BOC' ,   5200,'Pharmaceuticals','JASDAQ Standard'],
   ['2018-01-05',  2.810 ,   'BOC' ,   1980,'Pharmaceuticals','JASDAQ Standard'],
   ['2018-01-03',  5.199 ,   'CAT' ,   2000,'Real Estate','Mothers'],
  ['2018-01-06',  4.980 ,   'CAT' ,    450,'Real Estate','Mothers'],
  ['2018-01-07',  4.990 ,   'CAT' ,   3000,'Real Estate','Mothers']]
df = pd.DataFrame(data,columns =['date',  'price', 'ticker',  'volume', 'Sector','Market Division'])

我想要显示哪个市场部门有更多的库存，来自哪个部门。我尝试了下面的树状地图，但没有工作，我怎么做呢？

import plotly.express as px
import numpy as np

a=df.groupby(['Market Division','Sector']).count()

a["Exchange"] = "Exchange" # in order to have a single root node
fig = px.treemap(a, path=['Exchange', 'Market Division', 'Sector','ticker'], values='ticker')
fig.show()

Answer 1

您可以尝试使用stacked plots。下面是一个虚拟示例：

import matplotlib.pyplot as plt
labels = list(set([md for md in df['Market Division']]))
fig, ax = plt.subplots()
jasdaq = [3434, 5454, 45454] 
mothers = [35345, 64534, 43543]
ax.bar(labels, jasdaq[0], label='Pharmaceuticals')
ax.bar(labels, jasdaq[1], label='Communication')
ax.bar(labels, jasdaq[2], label='Real Estate')
ax.bar(labels, mothers[0], label='Pharmaceuticals')
ax.bar(labels, mothers[1], label='Communication')
ax.bar(labels, mothers[2], label='Real Estate')

ax.legend()
plt.show()

​

​

您需要首先计算每个Market分区的每个扇区，并替换jasdaq和mothers，以获得您想要的真实图。