Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Expand hypothesis testing to all linear models #4974

Draft
wants to merge 10,000 commits into
base: branch-23.02
Choose a base branch
from
Draft

Expand hypothesis testing to all linear models #4974

wants to merge 10,000 commits into from

Conversation

csadorf
Copy link
Contributor

@csadorf csadorf commented Nov 4, 2022 •
edited
Loading

Should be merged after #4952 and #4973.

daxiongshu and others added 30 commits February 22, 2022 18:29
@daxiongshu
Add get_params() to TargetEncoder (rapidsai#4588)
6cf64f6 
This PR adds `get_params()` member function to `TargetEncoder`. Hopefully it can resolve the issue rapidsai#4574

Authors:
  - Jiwei Liu (https://github.com/daxiongshu)

Approvers:
  - Victor Lafargue (https://github.com/viclafargue)
  - Dante Gama Dessavre (https://github.com/dantegd)

URL: rapidsai#4588
@hcho3
Upgrade Treelite to 2.3.0 (rapidsai#4590)
b8bc665 
The 2.3.0 version of Treelite incorporates the following improvements:
* GTIL optimization using multiple CPU threads (dmlc/treelite#353, dmlc/treelite#355, dmlc/treelite#357, dmlc/treelite#358, dmlc/treelite#362, dmlc/treelite#367)
* dmlc/treelite#365
* dmlc/treelite#366
* dmlc/treelite#368

Requires rapidsai/integration#436

Authors:
  - Philip Hyunsu Cho (https://github.com/hcho3)

Approvers:
  - William Hicks (https://github.com/wphicks)
  - Corey J. Nolet (https://github.com/cjnolet)
  - AJ Schmidt (https://github.com/ajschmidt8)

URL: rapidsai#4590
@robertmaynard
cuml now supports building non static treelite (rapidsai#4598)
5a18429 
Authors:
  - Robert Maynard (https://github.com/robertmaynard)

Approvers:
  - Dante Gama Dessavre (https://github.com/dantegd)

URL: rapidsai#4598
@cjnolet
Using Pyraft and automatically cloning when raft pin changes (rapidsa…
126c4ef 
…i#4593)

This PR depends on rapidsai/raft#520

Authors:
  - Corey J. Nolet (https://github.com/cjnolet)

Approvers:
  - AJ Schmidt (https://github.com/ajschmidt8)
  - Dante Gama Dessavre (https://github.com/dantegd)

URL: rapidsai#4593
@dantegd
Configurable libcuml++ per algorithm (rapidsai#4296)
c69606e 
Depends on rapidsai#4295 

PR allows `libcuml++` to be built with individual algorithms, or individual families of algorithms with the argument `CUML_ALGORITHMS`. It defaults to `ALL`, and can take multiple options like:

```
cmake .. -DCUML_ALGORITHMS="FIL;TREESHAP"
```

which will build a `libcuml++` only containing FIL and GPUTreeSHAP components. 

PR to update build documentation will follow up.

Authors:
  - Dante Gama Dessavre (https://github.com/dantegd)
  - Divye Gala (https://github.com/divyegala)

Approvers:
  - William Hicks (https://github.com/wphicks)
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#4296
@daxiongshu
Target Encoder with variance statistics (rapidsai#4483)
388a928 
This PR allows `TargetEncoder` to encode the `variance` of the target as requested by rapidsai#4440

Authors:
  - Jiwei Liu (https://github.com/daxiongshu)

Approvers:
  - Dante Gama Dessavre (https://github.com/dantegd)

URL: rapidsai#4483
@lowener
Remove classmethod of SimpleImputer (rapidsai#4439)
779d71b 
Closes rapidsai#4341.
The `classmethod` decorator seems to not be useful here and is blocking the serialization of SimpleImputer.

Authors:
  - Micka (https://github.com/lowener)

Approvers:
  - Victor Lafargue (https://github.com/viclafargue)
  - William Hicks (https://github.com/wphicks)

URL: rapidsai#4439
@RAMitchell
Fix OOM and cudaContext crash in C++ benchmarks (rapidsai#4594)
35d3c0a 
Fix rapidsai#4525 as well as a hard crash in c++ benchmarks due to some recent changes in raft.

Authors:
  - Rory Mitchell (https://github.com/RAMitchell)

Approvers:
  - Dante Gama Dessavre (https://github.com/dantegd)

URL: rapidsai#4594
@daxiongshu
Add a dummy argument deep to TargetEncoder.get_params() (rapidsai…
7d154b9 
…#4601)

This is the continuation of PR rapidsai#4588 to resolve issue rapidsai#4574

Authors:
  - Jiwei Liu (https://github.com/daxiongshu)

Approvers:
  - Dante Gama Dessavre (https://github.com/dantegd)

URL: rapidsai#4601
@viclafargue
Remove RAFT memory management (2/2) (rapidsai#4526)
d2099b8 
Authors:
  - Victor Lafargue (https://github.com/viclafargue)

Approvers:
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#4526
@lowener
Fitsne as default tsne method (rapidsai#4597)
701b62f 
Closes rapidsai#4566.

Authors:
  - Micka (https://github.com/lowener)

Approvers:
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#4597
@ajschmidt8
Merge 'branch-22.02' into 'branch-22.04'
a2c91a6
@Yosshi999
Fix document layouts in Parameters sections (rapidsai#4609)
b0cccb5 
like the figure below in https://docs.rapids.ai/api/cuml/stable/api.html#cuml.neighbors.NearestNeighbors
 
![image](https://user-images.githubusercontent.com/8027142/156071101-0a5bf96a-e073-4dea-8314-dfec733699c2.png)

Some captions are going into their heading.

Authors:
  - https://github.com/Yosshi999

Approvers:
  - Dante Gama Dessavre (https://github.com/dantegd)

URL: rapidsai#4609
@lowener
Add Complement Naive Bayes (rapidsai#4595)
2b0c200 
Closes rapidsai#1666.
The implementation of this variant is straightforward and matches sklearn.

Authors:
  - Micka (https://github.com/lowener)

Approvers:
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#4595
@lowener
XFail test_hinge_loss temporarily (rapidsai#4621)
7c3da85 
See rapidsai#3569. XFailing right now to unblock CI.

Authors:
  - Micka (https://github.com/lowener)

Approvers:
  - Dante Gama Dessavre (https://github.com/dantegd)

URL: rapidsai#4621
@RAMitchell
Kernel density estimation (rapidsai#4545)
b430b2e 
Using a brute force approach compared to sklearn's kd/ball tree. 

Todo:
- [x] Implement sample method
- [x] Sample weights
- [x] Evaluate which metrics are missing
- [x] Tests for sample
- [x] Docstrings

Authors:
  - Rory Mitchell (https://github.com/RAMitchell)
  - Corey J. Nolet (https://github.com/cjnolet)

Approvers:
  - Micka (https://github.com/lowener)
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#4545
@viclafargue
Use FAISS with RMM (rapidsai#4297)
4c1d671 
Answers rapidsai#2821

Authors:
  - Victor Lafargue (https://github.com/viclafargue)

Approvers:
  - Dante Gama Dessavre (https://github.com/dantegd)

URL: rapidsai#4297
@lowener
Add support for sample_weights in LinearRegression (rapidsai#4428)
fc94e5f 
Closes rapidsai#4031.
Scikit-learn is rescaling the data ([here](https://github.com/scikit-learn/scikit-learn/blob/0d378913be6d7e485b792ea36e9268be31ed52d0/sklearn/linear_model/_base.py#L313)) to take into account the sample_weight parameter.

Authors:
  - Micka (https://github.com/lowener)

Approvers:
  - Dante Gama Dessavre (https://github.com/dantegd)
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#4428
@vinaydes
Small fixes for certain test failures (rapidsai#4628)
c2594f1 
RAFT PR 513 changed meaning of probability for Bernoulli and Scaled Bernoulli distribution. This PR does corresponding change in cuML.

Authors:
  - Vinay Deshpande (https://github.com/vinaydes)

Approvers:
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#4628
@ajschmidt8
Add .github/ops-bot.yaml config file (rapidsai#4630)
072b32c 
In the near future, the [rapidsai/ops-bot](https://github.com/rapidsai/ops-bot) GitHub application that we use for GitHub automation will be enabled on all repositories in the `rapidsai` GitHub organization. Since not all features of the application are applicable to all repositories, this PR adds a new file, `.github/ops-bot.yaml`, which can configure which features are enabled per repository.

Authors:
  - AJ Schmidt (https://github.com/ajschmidt8)

Approvers:
  - Jake Awe (https://github.com/AyodeAwe)

URL: rapidsai#4630
@canonizer
Templatizing FIL types to add float64 support (rapidsai#4625)
05a9c16 
Templatize FIL types to add float64 support.

This is based on the work by @levsnv, specifically rapidsai#4569. This supersedes rapidsai#4569.

Authors:
  - Andy Adinets (https://github.com/canonizer)
  - Levs Dolgovs (https://github.com/levsnv)

Approvers:
  - Divye Gala (https://github.com/divyegala)

URL: rapidsai#4625
@dantegd
Add option to build RAFT artifacts statically into libcuml++ (rapidsa…
88ec85e 
…i#4633)

Add explicit option similar to FAISS and Treelite to be able to build a single `libcuml++` with all RAFT binary dependencies.

Authors:
  - Dante Gama Dessavre (https://github.com/dantegd)

Approvers:
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#4633
@galipremsagar
Upgrade dask & distributed minimum version (rapidsai#4632)
da89b78 
Rapids is upgrading to `2022.02.1` minimum version of dask. This PR updates those pinnings.

Authors:
  - GALI PREM SAGAR (https://github.com/galipremsagar)
  - Corey J. Nolet (https://github.com/cjnolet)

Approvers:
  - AJ Schmidt (https://github.com/ajschmidt8)
  - Dante Gama Dessavre (https://github.com/dantegd)
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#4632
@raydouglass
DOC
c04cee4
@cjnolet
Updates to consistency of MNMG PCA/TSVD solvers (docs + code consolid…
48120ce 
…ation) (rapidsai#4556)

Authors:
  - Corey J. Nolet (https://github.com/cjnolet)

Approvers:
  - Dante Gama Dessavre (https://github.com/dantegd)
  - Micka (https://github.com/lowener)

URL: rapidsai#4556
@GPUtester
Merge pull request rapidsai#4636 from rapidsai/branch-22.04
fbc1a50 
[gpuCI] Forward-merge branch-22.04 to branch-22.06 [skip gpuci]
@viclafargue
Remove RAFT MM includes (rapidsai#4637)
3f54aec 
Authors:
  - Victor Lafargue (https://github.com/viclafargue)

Approvers:
  - Corey J. Nolet (https://github.com/cjnolet)

URL: rapidsai#4637
@GPUtester
Merge pull request rapidsai#4640 from rapidsai/branch-22.04
f3afbac 
[gpuCI] Forward-merge branch-22.04 to branch-22.06 [skip gpuci]
@venkywonka
RF: fix stream bug causing performance regressions (rapidsai#4644)
baf1637 
This nanoPR fixes performance regression caused due to improper stream assignments to the decision trees.


Before fix:

 | sno |  algo | input | cu_time | cpu_time | cuml_acc | cpu_acc | speedup | n_samples | n_features | max_depth | n_estimators | n_bins | n_streams | n_jobs | n_classes | 
 | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | 
 | 0 | RandomForestClassifier | numpy | 32.635321855545044 | 0.0 | 0.99468 | 0.0 | 0.0 | 800000 | 64 | 8 | 500 | 128 | 4 | -1 | 2 | 
 | 1 | RandomForestClassifier | numpy | 40.36453413963318 | 0.0 | 0.994855 | 0.0 | 0.0 | 800000 | 64 | 10 | 500 | 128 | 4 | -1 | 2 | 
 | 2 | RandomForestClassifier | numpy | 61.35148477554321 | 0.0 | 0.99504 | 0.0 | 0.0 | 800000 | 64 | 16 | 500 | 128 | 4 | -1 | 2 | 

After fix:

| sno | algo | input | cu_time | cpu_time | cuml_acc | cpu_acc | speedup | n_samples | n_features | max_depth | n_estimators | n_bins | n_streams | n_jobs | n_classes
| --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- | --- |
0 | RandomForestClassifier | numpy | 28.637776374816895 | 0.0 | 0.99468 | 0.0 | 0.0 | 800000 | 64 | 8 | 500 | 128 | 4 | -1 | 2
1 | RandomForestClassifier | numpy | 34.11380743980408 | 0.0 | 0.994855 | 0.0 | 0.0 | 800000 | 64 | 10 | 500 | 128 | 4 | -1 | 2
2 | RandomForestClassifier | numpy | 47.153409481048584 | 0.0 | 0.99504 | 0.0 | 0.0 | 800000 | 64 | 16 | 500 | 128 | 4 | -1 | 2

Command run in `cuml/`
```
python python/cuml/run_benchmarks.py--num-rows 800000 --num-features 64 --skip-cpu --test-split 0.2 --cuml-param-sweep "n_bins=[128]" "n_streams=[4]" --cpu-param-sweep "n_jobs=[-1]" --param-sweep "max_depth=[8,10,16]" "n_estimators=[500]" --n-reps 1 --csv pool-2112-cls-800000.csv --dataset-param-sweep "n_classes=[2]" --dtype "fp32" --dataset classification -- RandomForestClassifier
```

Authors:
  - Venkat (https://github.com/venkywonka)

Approvers:
  - Dante Gama Dessavre (https://github.com/dantegd)

URL: rapidsai#4644
@GPUtester
Merge pull request rapidsai#4650 from rapidsai/branch-22.04
68e70c2 
[gpuCI] Forward-merge branch-22.04 to branch-22.06 [skip gpuci]
csadorf added 14 commits December 5, 2022 06:15
@csadorf
Hypothesize test_weighted_linear_regression test.
63fcd12
@csadorf
Hypothesize test_ridge_regression_model_default test.
55c91ad
@csadorf
Hypothesize test_ridge_regression_model test.
73960b0
@csadorf
Mark test_ridge_regression_model_default as xfail.
cd3db50
@csadorf
Hypothesize test_weighted_ridge test.
28fe66d
@csadorf
Implement standard_classification_datasets hypothesis strategy.
a0af4db
@csadorf
Hypothesize test_logistic_regression.
ac58201
@csadorf
Hypothesize test_logistic_regression_unscaled test.
8b97769
@csadorf
Hypothesize test_logistic_regression_model_default test.
db0017e
@csadorf
Hypothesize test_logistic_regression_model_digits test.
9b59a3d
@csadorf
Hypothesize test_logistic_regression_sparse_only test.
ce341af
@csadorf
Hypothesize test_logistic_regression_decision_function test.
f7dda37
@csadorf
Hypothesize test_logistic_regression_predict_proba test.
c01b867
@csadorf
Hypothesize test_ridge_predict_convert_dtype test.
2fbe4a5
@csadorf csadorf force-pushed the fea-expand-hypothesis-testing-to-all-linear-models branch from 6849d71 to a09b558 Compare December 6, 2022 12:12
@csadorf csadorf mentioned this pull request Dec 6, 2022
Merged
@csadorf
Hypothesize test_logistic_predict_convert_dtype test.
dd6283d
@csadorf
Better default of n_informative for standard_classification_datasets.
18b27c4 
- The strategy chooses the n_informative default strategy more smartly
to satisfy the inequality assumption between the numnber of classes and
clusters per class and number of informative features.
- The strategy tries to prevent a more informative error message in case
that the assumption cannot be met with the given parameters arguments.
@csadorf
Hypothesize test_logistic_regression_weighting test.
c3e1105
@csadorf
Hypothesize test_linear_models_set_params test.
2b9a447
@csadorf
Hypothesize test_elasticnet_solvers_eq test.
e0dda4a
@csadorf
Remove obsolete utility functions and imports.
b8baffe
@csadorf csadorf force-pushed the fea-expand-hypothesis-testing-to-all-linear-models branch from c3149ad to b8baffe Compare December 6, 2022 20:20
@csadorf csadorf changed the title Fea expand hypothesis testing to all linear models Expand hypothesis testing to all linear models Dec 6, 2022
@csadorf csadorf added 0 - Blocked Cannot progress due to external reasons and removed 2 - In Progress Currenty a work in progress labels Dec 15, 2022
@csadorf
Copy link
Contributor Author

csadorf commented Dec 15, 2022

We should merge #5065 and address #4963 before moving forward with this.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
0 - Blocked Cannot progress due to external reasons Cython / Python Cython or Python issue improvement Improvement / enhancement to an existing function non-breaking Non-breaking change
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.