You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
We now have all the scaffolding to add binary hybrid handlers, e.g. +, -, == at least handling scalar values, so that we can hybridify something like this:
sum( x < 3 )
We know how to handle sum, but since we can't hybrid x<3 the hybrid version of sum is not used.
The text was updated successfully, but these errors were encountered:
@romainfrancois do think you can tackle this with the other hybrid evaluator changes? Or should we push off to a future version? I'm aiming to get dplyr 0.3 to CRAN by mid-September.
For this to really be useful, we would have to specialize precisely for each case, i.e. at the point where we see the expression sum( x < 3 ) we would instantiate a class like Sum_LowerThan.
Otherwise, within hybrid eval, we would first substitute x < 3 to some logical vector and then sum that which demands more memory and runtime .
We now have all the scaffolding to add binary hybrid handlers, e.g.
+
,-
,==
at least handling scalar values, so that we can hybridify something like this:sum( x < 3 )
We know how to handle
sum
, but since we can't hybridx<3
the hybrid version ofsum
is not used.The text was updated successfully, but these errors were encountered: