1
$\begingroup$

I was trying to calculate the frequency of disease_present (yes) when smoking status is y (yes) for each group (A, B, C, D)

The formula is = Number of samples with disease_present (yes) / No of samples with smoking status 'y' * 100

df

Sample_code Group   smoking disease_present
1   A   y   yes
2   B   n   
3   C   n   yes
4   D   y   
5   B   n   yes
6   A   y   yes
7   C   n   
8   D   y   
9   A   n   yes
10  B   y   
11  D   n   yes
12  C   n   
13  A   y   yes
14  B   y   yes

I tried this

df.groupby("Group").count()

This gave the same number in every category

  1. Pandas by default turns empty cells into NaN
  2. I tried to replace NaN with empty strings with df1 = df.fillna(""). I still got the same results
$\endgroup$

1 Answer 1

2
$\begingroup$

I am not able to reproduce the issue.

However I think this is a crosstab problem:

>>> new_df = df.fillna("no")
>>> new_df
# prints:
    Sample_code Group smoking disease_present
0             1     A       y             yes
1             2     B       n              no
2             3     C       n             yes
3             4     D       y              no
4             5     B       n             yes
5             6     A       y             yes
6             7     C       n              no
7             8     D       y              no
8             9     A       n             yes
9            10     B       y              no
10           11     D       n             yes
11           12     C       n              no
12           13     A       y             yes
13           14     B       y             yes

>>> pd.crosstab([new_df.disease_present, new_df.smoking], new_df.Group)
# prints
Group                    A  B  C  D
disease_present smoking            
no              n        0  1  2  0
                y        0  1  0  2
yes             n        1  1  1  1
                y        3  1  0  0
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.