NicolasHug
Follow
Create your own GitHub profile
Sign up for your own profile on GitHub, the best place to host code, manage projects, and build software alongside 50 million developers.
Sign up
Pinned
937 contributions in the last year
Activity overview
Contributed to
scikit-learn/scikit-learn,
scikit-learn/enhancement_proposals,
NicolasHug/Surprise
and 5 other
repositories
Contribution activity
September 2020
Created a pull request in scikit-learn/scikit-learn that received 15 comments
[MRG] MNT Initialize histograms in parallel and don't call np.zero in Hist-GBDT
To keep things more unitary, this PR extracts out the histogram initialization from #18242 which was already proposed in #14392 Memory usage is the…
+16
−2
•
15
comments
- [MRG] Forbid splitters that yield different splits in SequentialFeatureSelection
- [MRG] MNT Use a pytest-like context manager in estimator_checks.py
- [MRG] MNT Don't validate error messages when strict mode is off in check_estimator
- [MRG] Fix accept uint8 X in hist-gbdt.predict
- [MRG] Support unknown_value=np.nan in OrdinalEncoder
- [MRG] DOC Good practices and pitfalls for random_state
- [MRG] Add stratify option for learning_curve
- [MRG] DOC Good practices and pitfalls for random_state
- DOC specify default of error_score in cross_validate
- DOC Minor docstring update
- DOC Add common pitfalls: data leakage
- DOC Fix step 5 under "Contributing Code" to include build tools installation
- FEA Allow passing custom covariance estimators to LDA
- [MRG] Fix accept uint8 X in hist-gbdt.predict
- ENH Adds Categorical support for HistGradientBoosting
- DOC Add enable_successive_halving module to the API Reference
- CLN Refactors find binning threshold
- BENCH threading scalability of Hist Gradient Boosting
- DOC Fix x-axis label positions for SH iterations example
- Remove preprocessing functions from docs.
- DOC Nitpicks in the API Reference
- [MRG] MNT Initialize histograms in parallel and don't call np.zero in Hist-GBDT
- ENH Break cyclic references in Histogram GBRT
- [MRG] DEP deprecate "mae" criterion in GadientBoosting estimators
- ENH Greatly reduces memory usage of histogram gradient boosting
- [MRG] Successive halving for faster parameter search
- CI Take comment only assigns issue if there are not assignees
- MNT Use version less than 3.0.0 for lightgbm
Created an issue in scikit-learn/enhancement_proposals that received 2 comments
Questions and comments about SLEP006 -- Sample props
I've made a pass over the SLEP. It is overall very clear and the different cases really help. Thanks @jnothman and @adrinjalali for your efforts. H…
2
comments