python - Efficient creation of a new time series from two irregular time series with Pandas -
i have 2 time irregular series, , b, want create new one. resulting series should have same index values should based on rolling sum on time window of values in b, in relation indices in a.
for example: a
2011-01-27 10:21:43 0 2011-01-27 10:43:29 0 2011-01-27 19:39:39 0 2011-01-27 19:55:55 0 2011-01-27 19:58:25 0 2011-01-28 15:31:58 0 2011-01-28 16:27:13 0
b
2011-01-27 10:20:29 0 2011-01-27 18:31:23 1 2011-01-27 18:45:25 1 2011-01-27 18:57:22 1 2011-01-27 19:15:25 0
desired result using 1 hour window:
2011-01-27 10:21:43 0 2011-01-27 10:43:29 0 2011-01-27 19:39:39 2 2011-01-27 19:56:55 1 2011-01-27 19:58:25 1 2011-01-28 15:31:58 0 2011-01-28 16:27:13 0
currently looping through indices of , performing sum on b using b[t-hour():t].sum(). seems inefficient. suggestions?
Comments
Post a Comment