Martin Alfaro

I just also noticed that adding an intercept in this way crashes `predict` if there are missing values (it works fine if there are no missing values). The MWE is:...

There's an operation I perform a lot. I'm adding here because I think it's related. Maybe this is worth considering now that the API is under development. Alternatively, it'd be...

Yes, that was my point. I use `combine` + `leftjoin!` to avoid having missing values, and because methods like the second example are less general and only applicable to specific...

Is the problem here the syntax or the speed? If it's the former, I'm not sure if there's any benefit from notation like this ```julia df[(:name => "Sally", :age =>...

If it's for speed and correctness, I like the idea. Maybe we could think more about the API? the API and name of `lookup` doesn't convince me. I'm thinking of...

Just my two cents as a user, I think that the problem is then more related to the documentation. I'd expect the use of groupby for subsetting a dataframe [here](https://dataframes.juliadata.org/stable/man/working_with_dataframes/#Taking-a-Subset)....

One more comment about the documentation. The section API and in particular the part [you mention](https://dataframes.juliadata.org/stable/lib/indexing/#Indexing-GroupedDataFrames) has been always confusing to me. Some of the reasons. 1. Since it's not...

A small question about speed because I kept thinking about the use of `groupby` for looking up values. Testing that option, it always gives me the slowest option. ```julia using...

>Why does this error? See the first post of this discussion. That was the original proposal. The idea is that you need a tuple, such that `iris_gdf[("Iris-virginica",)]` I misinterpreted the...

My two cents about why I wouldn't recommend adding a new method: 1. The operation can be implemented in other ways already. The more methods to implement the same feature,...