0.4.0 #355
Merged
0.4.0 #355
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
- shell script edit - baselearner typo fix - set_args_calc cleaning due to amadeus process_hms refactoring - NOTE: set_args_calc must be run after loading amadeus
- feature_raw_download: duplicated arguments `dataset_name` removed - set_args_download: data_files append reverted
- previous push changed set_args_calc; corrected
Merge main into pipeline-compact
- per amadeus update
- Per download_modis logic, date arguments will be in the same year
- Skeleton metalearner per the new strategy - Previous meta learners are dropped - set_args_download: narr edit
Changes during OOO July 2024
- Refactored fit_* by abstracting reusable parts - fit_* functions return the tuned hyperparameters and their prediction values from the entire data - Renaming cv index generation functions into generate_* - make_subdata will only return row indices - README.md for targets is updated - Bump version and roxygen2 documentations rebuilt
- fit_base_learner: simplified branching-friendly interface for base learner - inject_match: inject matching arguments for making evaluation fail-safe - generate_* functions now return holdout group indices or list of row indices - fit_base_[modelname] will be removed from the codebase - Documentation fix - vis_spt_set: added cex argument
- Unnecessary lines are removed from targets_baselearner.R
- hard-coded search path is parametrized - targets_calculate.R is modified
added 21 commits
August 16, 2024 13:31
- bayes to grid (random grid search with size 10)
- nthreads is restored in fit_base_learner for multithreading in tuning process
- "lightgbm" to "lgb" - _targets.R: max future globals size set 50GB
- "bonsai", which is required to define a parsnip model with lightgbm, is added to `tar_option_set()`
- dials approach is not compatible with the parsnip+tune model declaration - Added a list of data.frame target for candidate hyperparameters that match what were declared to be tuned, pick 10 rows for grid search
- R-lightgbm seemingly leverages OpenCL in some parts and all CPU threads in others
- README.md: clear on GPU-enabled packages and warnings about xgboost version - typos in grid definition (c instead of seq) - fit_base_learner() and fit_base_tune(): returns performance metrics from all tuned models for manuscript and future reporting - save_pred is deactivated in tune_*() options
- Elaborate documentations - Added some tests replacing previous meta learner tests - Drop soft-deprecated base learner functions (moved to reserved.R)
- .lintr: added entire /tests folder - test errors and warnings are fixed - lint
- DESCRIPTION: added tibble in Imports - Vignettes built; aqs_download is moved to /tools - vignettes folder is removed from .Rbuildignore - Lint - pkgdown.yaml gets remote package installation - /doc removed from .gitignore
- vignette file reference to tigris::counties call - Missing arguments in Rd were fixed - fit_base_learner: dropped vfold, folds will get integer number of folds
There was a problem hiding this comment.
Have you added a workflow that creates the full coverage report by line and is uploaded somewhere on GitHub? If so, that is awesome!
kyle-messier
approved these changes
Aug 19, 2024
|
I tried to host this file directly on pkgdown website, but it was so confusing to set the correct path in pkgdown 😅, so I uploaded the report file as an artifact upon passing all tests. You can click the link in artifact to download the html file.
…________________________________
From: {SET}group ***@***.***>
Sent: Monday, August 19, 2024 17:04
To: NIEHS/beethoven ***@***.***>
Cc: Insang Song ***@***.***>; Author ***@***.***>
Subject: Re: [NIEHS/beethoven] 0.4.0 (PR #355)
@kyle-messier commented on this pull request.
________________________________
On .github/workflows/test-coverage.yaml<#355 (comment)>:
Have you added a workflow that creates the full coverage report by line and is uploaded somewhere on GitHub? If so, that is awesome!
—
Reply to this email directly, view it on GitHub<#355 (review)>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/AGCFCUQEW6ULTFC5FVD2M7DZSJMWVAVCNFSM6AAAAABMV2S32SVHI2DSMVQWIX3LMV43YUDVNRWFEZLROVSXG5CSMV3GSZLXHMZDENBWGQ2TONBZGA>.
You are receiving this because you authored the thread.Message ID: ***@***.***>
|
mitchellmanware
added a commit
that referenced
this pull request
Aug 23, 2024
Merge pull request #355 from NIEHS/pipeline-compact
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
No description provided.