Thomas J. Fan
Thomas J. Fan
When the numerical data is scaled the one hot encoder works pretty well: ```py from sklearn.compose import ColumnTransformer from sklearn.pipeline import Pipeline from sklearn.preprocessing import OneHotEncoder from sklearn.preprocessing import StandardScaler...
#### Reference Issues/PRs Related to https://github.com/scikit-learn/scikit-learn/pull/23865 #### What does this implement/fix? Explain your changes. As noted in https://github.com/scikit-learn/scikit-learn/pull/23865#discussion_r941346271, I agree that using `DTYPE_t` and `ITYPE_t` is a bit confusing and...
#### Reference Issues/PRs Towards https://github.com/scikit-learn/scikit-learn/issues/5870 #### What does this implement/fix? Explain your changes. This PR adds missing values for trees for the dense case and the best splitter. I think...
### Background / Objective [Docstrings](https://www.python.org/dev/peps/pep-0257/#what-is-a-docstring) in Python are string literals that occur as the first statement in a module, function, class, or method definition. These are *some* of the characteristics...
#### Reference Issues/PRs Closes #24105 #### What does this implement/fix? Explain your changes. The most recent change to `SpectralEmbedding` came from: https://github.com/scikit-learn/scikit-learn/pull/23210. I suspect it has to do with setting...
#### Reference Issues/PRs Fixes #19051 #### What does this implement/fix? Explain your changes. I think `text_kw` is okay to add since `text` is called frequently to place values into the...
I am proposing allowing a `kwargs` to be passed to the parallel_backend to configure the parallel call: ```python from joblib import Parallel, delayed, parallel_backend from math import sqrt with parallel_backend("loky",...
I had a recent conversation with a user that tried to use `DistributedDataParallel` with `skorch`'s early stopping and this would cause the process to hang. My guess is that since...
Continuing from PR https://github.com/dnouri/skorch/pull/360 for `TrainEndCheckpoint`. When one presses Ctrl+C in the middle of an epoch, the latest epoch would most likely be in an incomplete state. When the checkpoint...
Two API questions: 1. Should `history` be in `__init__`? This seems to be used for continue training. One would need to pass in an initialized `module`, `criterion`, `optimizer`, `history`, and...