Suppose that I want to create a new column of integers, default them all to
"not set" (in other words, NA), and then loop and initialize some of them
later.
I can't just `df[:C] = NA` because then I'll have a column that's an
Array{NA,1}...
So maybe I've got to do something like:
df[:C] = fill!(Array(Any, size(df,1)), NA)
But then I'm sort of breaking the DataFrame structure (as I understand it).
Underneath, the DataFrame is suppose to be a nicely typed set of column
arrays, with a separate set of columns that contain values that indicate
when something is missing. What I just produced is a column with a very
generic type where all values are set and some just happen to be the
special NA value.
Is there a better way to do this?
On Friday, May 16, 2014 7:10:06 AM UTC-4, Jason Solack wrote:
>
> Thank you!
