order could utilize existing index #3023

Activity

1. jangorecki

   added a commit that references this issue on Aug 30, 2018

   in dev index optimize for order call

   240c936
2. jangorecki

   changed the title [-]order could utilize existing index[/-] [+]`order` could utilize existing index[/+] on Aug 30, 2018
3. jangorecki

   added

   performance

   on Aug 30, 2018
4. jangorecki

   self-assigned this

   on Apr 14, 2020
5. jangorecki commented on Apr 19, 2020

   

   jangorecki

   on Apr 19, 2020

   Last edited by jangorecki

   MemberAuthor

   More actions

   This issue could be automatically handled by #4386 but the problem is that forder strips out all attributes before even calling forderv. This happens because when we call

   d = data.table(x=2:1)
   d[order(x, na.last=FALSE)]

   We technically call forder on integer vector x, not on a d data.table. To not duplicate the work, to address that, it is best to rely on #4386 and we have to make better input massaging when preparing arguments for forderv, retaining key/index attributes.
   According to ideas in #4346 the proper way to address that would be to optimize from [.data.table directly to forderv, and not to forder.
6. jangorecki commented on May 11, 2020

   

   jangorecki

   on May 11, 2020

   MemberAuthor

   More actions

   It should be possible to optimize as well d[order(x)] in cases where x does not have NAs, which can be a cheap to check if index exists.
7. jangorecki

   linked a pull request that will close this issueorder optimizes to forderv, uses key or index, forder more consistent to base order #4458on May 18, 2020