Rory Finnegan
Rory Finnegan
Hmm, I believe this was introduced to avoid having LOCF silently fail to impute any values. Perhaps we should support a flag or something... If you're positive that you don't...
Nice, thanks for all the debugging details! Hmm, I'm not sure if that's exactly a bug. If you try and interpolate among only `missing` values then I'd expect that to...
I'm afraid I was unable to reproduce your example on my local mac (1.6.1) or in a `julia:1.6.4-buster` docker container, using a range of 6-12 threads. Would you mind calling...
Actually, I don't think this would be breaking because the `Tuple` constructor would remain. Bumping this to the 1.0 milestone.
> I haven't really looked into why impossible It's impossible because to guarantee safety, you'd need to: 1. Manually serialize to a static set of primitive types that will never...
Another issue stems from the fact that we can't guarantee that `Impute.locf!` won't copy the data. This works for dataframes, but that isn't a consistent behaviour. I'm not sure if...
How so? Isn't the point that you don't want to copy each group?
Ahh, oops, I missed the impact of the append. Yeah, that make sense now.
Okay, I'll try giving that a shot... though I'm guessing we'll need more tests cases for different types of tables then.
That's because it isn't a current feature. Currently, if you want to work on a subset of the table then you should select those columns first.