[FIXED] How to manipulate 2 column data frame in Jupyter notebooks Python

Issue

I have a dataframe that I am trying to manipulate to show the difference in severity for accidents that occur in dark vs light conditions.

This is the df that has 200k entries.

SEVERITYCODE    LIGHTCOND
0   Injury  Light
1   Damage  Dark
2   Damage  Light
3   Damage  Light
4   Injury  Light
5   Damage  Light
6   Damage  Light
7   Injury  Light
8   Damage  Light
9   Injury  Light
10  Damage  Light
11  Damage  Light
12  Damage  Dark
13  Damage  Dark
14  Injury  Dark
15  Damage  Dark
16  Injury  Light
17  Damage  Light
18  Injury  Light
19  Damage  Dark
20  Injury  Dark

I need to be able to get this data in to the df to see something like this where the number of occurrences of dark-injury are in the upper left hand box, dark and damage are in the upper right hand box, and so on.

       Injury   Damage
Dark:   10023   1132
Light:  1234    98474

How do I make Python count across columns like this? I wasn't sure if the data in the included picture is required to help me out or not.1

I then want to make it into a stacked bar graph for easy visualization which I think I can manage through other tutorials.

Thanks

Solution

(
    df.groupby(['LIGHTCOND', 'SEVERITYCODE']) # create a groubpy object
    .size() # aggregate by counting the rows in each group
    .unstack() # move the inner-most index level to columns, i.e. 'SEVERITYCODE'
)

Answered By - RichieV