Plotting a scatter plot of a binary outcome variable over time

I have a large data frame consisting of extracted forum posts that looks something like this

| user_name | thread_name || post_text | time_of_post || date_joined | binary_target |
| -------- | -------- || -------- | -------- || -------- | -------- |
| BoxCutter |It's Okay When We Do It. || ..... | 2022-08-09 19:39:00 || 2022-05-26 | 1 |
| Docket_33 |It's Okay When We Do It. || ..... | 2022-08-09 19:54:00 || 2022-06-10 | 1 |
| Hearmeout |It's Okay When We Do It. || ..... | 2022-08-09 19:58:00 || 2021-10-07 | 0 |

I have run a binary classifier on the post_text column, giving instances of 1 or 0 in the binary_target column. I wish to plot, for each date present in the data frame (time_of_post), the number of posts classified as a 1. However, since the total number of posts for each day varies, I would like to first calculate the total number of 1s for each date as a percentage of the total number of posts for each date.

The output I desire will be a scatter plot, with date on the x axis, and the y axis being 'instances of 1s as a % of total posts'.

I am able to easily obtain the value counts of the binary targets using:

df_combined.groupby('date_joined')['binary_target'].value_counts()

Though I am struggling with calculating them in percentage terms.

You can get the number of 1s by taking the sum of the observations on each date, then divide that by the total count of observations on each date. You can get both of those using groupby and agg.

df_agg = df_combined.groupby('date_joined')['binary_target'].agg(['sum', 'count'])
print(df_agg)

From there you can create a new column to compute the proportion of 1s on each date.

df_agg['prop'] = df_agg['sum'] / df_agg['count']
print(df_agg)