绘制条形图和 Groupby 关键字

Question

I have a pandas data frame similar to:

Hospital                           2009-10  2010-11
Llandudno General Hospital         43       54
Dolgellau District Hospital        57       58
Deeside Community Hospital         120      140
Haverfordwest Mental Health Unit   34       30

and I want to make a bar plot of the different types of hospitals by keyword ie 'Mental Health', 'District'. Grouping all the 'Mental Health' hospitals together, Grouping all the 'District' Hospitals together etc.

Here is my code so far:

bedsByType = df[ ['Hospital', '2009-10', '2010-11'] ].groupby(['Mental Health', 'General' , 'Community','District'])

summedAndSortedBedsByType = bedsByType.sum().sort_values( '2009-10')

summedAndSortedBedsByType.plot.barh(figsize=(25,15), title='Different Types of Hospitals')

Answer 1

It is not really specified in your question, how you determine your groups. I assume there exists as list for the categories. Then you can create your graph for instance like this:

import pandas as pd
from matplotlib import pyplot as plt 

#sample df

                                Hospital  2009-10  2010-11
0             Llandudno General Hospital       43       54
1            Dolgellau District Hospital       57       58
2             Deeside Community Hospital      120      140
3       Haverfordwest Mental Health Unit       34       30
4  Morelake General Mental Health Clinic       37       39
5       Manderlay Mental Health Hospital       17       29
6             Cumbria Community Hospital       28       25
7                       Mayfair Hospital       17       19
8             New Kent District Hospital       14       17
#define categories in a list
groups = ["Mental Health", "General", "Community", "District"]
#create pattern for grouping
pattern = "|".join(groups)
#create new column with categories, if nothing applies use a fill value
df["type"] = df["Hospital"].str.extract("({})".format(pattern), expand = False).fillna("N/A")
#sum bed numbers for each category
df1 = df.groupby("type")["2009-10", "2010-11"].sum()
#create bar chart
df1.plot.barh(title = "Beds by hospital type")
plt.show()

Output: