I am a multi year data.table user and today I've seen 3 times slower performance of data.table time based rolling aggregation compared with pandas rolling aggregation. Please advise if the behavior is expected and if here is a correct place to post reproducible examples. Thanks.
for data.table time based rolling I use non-equip join syntax. Pandas rolling I use the rolling() function.
I am a multi year data.table user and today I've seen 3 times slower performance of data.table time based rolling aggregation compared with pandas rolling aggregation. Please advise if the behavior is expected and if here is a correct place to post reproducible examples. Thanks.
for data.table time based rolling I use non-equip join syntax. Pandas rolling I use the rolling() function.