Closed piever closed 7 years ago
Indeed, the f.instance
on that line wasn't a very clean approach. I guess that's what happens when you do something without knowing why it works...
I have a simple fix, but I need to check that performance doesn't suffer first.
See https://github.com/JuliaStats/NullableArrays.jl/pull/177. Unfortunately, inference fails on Julia 0.5 and I couldn't find a workaround (at least it works on 0.6).
After #175 I can pass nullable arrays with different
eltype
tobroadcast
but it still doesn't seem to work on type constructors.Example:
On the other hand, this works (but it fails the inference part as the output should be of type
NullableArrays.NullableArray{SurvEvent,1}
):