Search code examples
pythonpandasgroup-byjupyter-notebook

Figure Percentage by Group in Pandas

My data looks like the below

| Position | Enrolled | Retained
| -------- | -------- |
| FT       | 1        |  1
| PT       | 1        |  0
| S        | 1        |  1
| FT       | 1        |  0
| S        | 1        |  0
| PT       | 1        |  1

I want to calculate the percent retained (Retained/Enrolled) by group.

I'd like my output to look like

FT .50
PT .50
S  .50

I tried the following:

df2.groupby(by=['Position']).(df2['Enrolled'].sum()/df2['Retained'].sum())

But that's not the correct syntax to accomplish this.

Solution

  • If you use pd.DataFrame.eval, you can get pretty close (syntactically) to your proposed solution ;)

    out = df2.groupby("Position").sum().eval("Retained / Enrolled")

    out:

    Position
FT    0.5
PT    0.5
S     0.5
dtype: float64