Accept boosters as model inputs?

[lbittarello]

Model currently requires the path to a model file. I was wondering if it'd make sense to also accept a booster. We could call to_string and save it as a temporary file or just work with the string representation directly. It'd make users' life (a little) easier.

[siboehm]

I can see how this would make using lleaves slightly easier, but it's a single line for the user (saving the output of Booster.model_to_string() to a tmpfile and passing that to lleaves) and it just complicates the API. We could write a code example of how to save to a tmpfile, but I don't see a big reason to put this functionality into lleaves.

[lorey]

Also, there's io.StringIO('foo') to get a file handle from a string thus omitting the tmp file: https://docs.python.org/3/library/io.html#text-i-o

[lbittarello]

io.StringIO('foo') won't work: TypeError: stat: path should be string, bytes, os.PathLike or integer, not StringIO.

We're currently doing:

import tempfile
from pathlib import Path

path = Path(tempfile.mkdtemp(), "booster.txt")

with open(path, "w+") as f:
    f.write(self.booster_.model_to_string())

lleaves_model_ = lleaves.Model(path)

which is not awful, but it isn't as practical as it could have been, especially as lleaves should only really need the string representation (I think) and saving to disk is just wasteful. 🤷