How to set the column of the last row of each group to the previous row's column

I have a dataframe df, that has columns id, time, start_quantity, end_quantity. For each id and time, the start_quantity is equal to the end_quantity at the previous time.

Here's an example:

id   time       start_quantity      end_quantity
0    1          0                   10
0    2          10                  15
.....
23   1          55                  87       
23   2          87                  90
.....

There's a degenerate case in df, where the last row of each id (note that dataframe is pregrouped by id and within each id group its sorted in ascending order based on time), has an incorrect start_quantity. The last row for each id is always time = 10.

For each of these rows, I would like to make the correction, but when I tried

df.loc[df[df.time == 10], "start_quantity"] = df.loc[df[df.time == 9], "end_quantity"]

It makes the start_quantity for those rows NaN.

Shift the values in the end quantity per id, then use boolean indexing with loc to update the values where time == 10

s = df.groupby('id')['end_quantity'].shift()
df.loc[df['time'] == 10, 'start_quantity'] = s

Alternative approach:

df.loc[df['time'] == 10, 'start_quantity'] = df.loc[df['time'] == 9, 'end_quantity'].tolist()