Matplotlib graphic on a particular dataframe from groupby-CodePudding

i want to have a graphic with x axis for TOT Population, y axis for Years and two lines one for Hispanic and one for not Hispanic. The dataframe is that:

                                  ID Race   ID Ethnicity    ID Year Hispanic Population Moe
Ethnicity          TOT Population               
Hispanic or Latino  9825          4.0           1.0          2013.0         2345.0
                    12234         4.0           1.0          2014.0         2660.0
                    12437         4.0           1.0          2018.0         2429.0
                    13502         4.0           1.0          2016.0         3254.0
                    14025         4.0           1.0          2019.0         2644.0
... ... ... ... ... ...
Not Hispanic or Latino  
                    14616636      0.0           0.0          2017.0         7788.0
                    14725729      0.0           0.0          2016.0         8629.0
                    14815122      0.0           0.0          2015.0         7888.0
                    14849129      0.0           0.0          2014.0         7495.0
                    14884539      0.0           0.0          2013.0         6586.0

I got this dataframe from a groupby of Ethnicity and TOT Poplation. SOmeone can help me to make real a matplotlib? Thank you!

CodePudding user response：

I believe there are two parts to your question. First is to move the grouped data to a format that maatplotlib would understand (basically flatten the table) and second to plot (lines) the two lines in one graph.

The initial data:

>> df
                                     ID Race    ID Ethnicity    ID Year  Hispanic...
Ethnicity          TOT Population               
Hispanic or Latino           9825       4               1        2013       2345
                             12234      4               1        2014       2660
                             12437      4               1        2018       2429
                             13502      4               1        2016       3254
                             14025      4               1        2019       2644
Not Hispanic or Latino      14616636    0               0        2017       7788
                            14725729    0               0        2016       8629
                            14815122    0               0        2015       7888
                            14849129    0               0        2014       7495
                            14884539    0               0        2013       6586

First, use reset_index to flatten the table

>> df2 = df.reset_index()
>> df2
   Ethnicity    TOT Population  ID Race ID Ethnicity    ID Year Hispanic Population Moe
0   Hispanic or Latino  9825    4   1   2013    2345
1   Hispanic or Latino  12234   4   1   2014    2660
2   Hispanic or Latino  12437   4   1   2018    2429
3   Hispanic or Latino  13502   4   1   2016    3254
4   Hispanic or Latino  14025   4   1   2019    2644
5   Not Hispanic or Latino  14616636    0   0   2017    7788
6   Not Hispanic or Latino  14725729    0   0   2016    8629
7   Not Hispanic or Latino  14815122    0   0   2015    7888
8   Not Hispanic or Latino  14849129    0   0   2014    7495
9   Not Hispanic or Latino  14884539    0   0   2013    6586

You then plot the line graph.

import matplotlib.pyplot as plt
plt.figure(figsize=(20, 5))

plt.plot(df2['TOT Population'][df2['Ethnicity'] == 'Hispanic or Latino'], df2['ID Year'][df2['Ethnicity'] == 'Hispanic or Latino'])
plt.plot(df2['TOT Population'][df2['Ethnicity'] == 'Not Hispanic or Latino'], df2['ID Year'][df2['Ethnicity'] == 'Not Hispanic or Latino'], '-.')
plt.ticklabel_format(style='plain')
plt.xlabel("TOT Population")
plt.ylabel("Year")
plt.title('My plot')

Your graph will look like this. You can change it further as you need. Note that the Hispanic population is rather small compared to the non-hispanic population. So, the graph was made rather wide. You can plot just one group and see the ups and downs better.

Output graph