Closed cortner closed 1 year ago
CC @tjjarvinen -- I believe this would take you no time at all, and fits with your general work on the python interface.
Ok, so the idea is that you want Python SkLearn to be an option for fitting, but normally you would use Julia Optim.jl?
For what I looked you did not use ScikitLearn.jl, but instead called SkLearn from Python with PyCall. This will make it easy to convert to PythonCall.
correct. At the moment, some people still use this, but over time there will likely be fewer and fewer as Chuck's code is improving.
I’m fine with this, but note that PyCall is a now a Requires dependency, in case you have to undo that.
On Mon, 19 Jun 2023 at 18:32, Christoph Ortner @.***> wrote:
correct. At the moment, some people still use this, but over time there will likely be fewer and fewer as Chuck's code is improving.
— Reply to this email directly, view it on GitHub https://github.com/ACEsuit/ACEfit.jl/issues/51#issuecomment-1597466564, or unsubscribe https://github.com/notifications/unsubscribe-auth/ACXHHTS3H7FFWZ7CNR6CMELXMB5IHANCNFSM6AAAAAAZMFBJVU . You are receiving this because you are subscribed to this thread.Message ID: @.***>
Thanks. I'll check whether PyCall can be removed from [dep]
before merging and tagging
I was unable to remove the PyCall and don't have time to figure this out, so I removed Requires instead, sorry about that.
In my opinnion we can close this issue, as #53 has implemented this.
Makes sense - closing