Ecosyste.ms: Issues
An open API service for providing issue and pull request metadata for open source projects.
GitHub / fmohr/naiveautoml issues and pull requests
#78 - added runtime forecast for evaluators (currently only invoked for SVMs)
Pull Request -
State: closed - Opened by fmohr 14 days ago
#77 - Timeout doesn't work properly on SVMs sometimes
Issue -
State: closed - Opened by fmohr 14 days ago
#76 - Add support for multi-label classification
Issue -
State: open - Opened by mwever about 2 months ago
#75 - Problems when ruuning with the AMLB and shorter time constraints
Issue -
State: open - Opened by israel-cj 3 months ago
#74 - Including description and example on how to exclude steps or components
Issue -
State: open - Opened by tilly111 3 months ago
#73 - NAML does currently not allow to set the training set size in the evaluators
Issue -
State: open - Opened by fmohr 3 months ago
#72 - allow to use name of scoring function
Pull Request -
State: closed - Opened by fmohr 4 months ago
#71 - naive automl should be able to receive a scorer object if it has a name
Issue -
State: closed - Opened by fmohr 4 months ago
#70 - fixed issues with configspace
Pull Request -
State: closed - Opened by fmohr 4 months ago
#69 - Naive AutoML generates tons of warning because of a change in the ConfigSpace library
Issue -
State: closed - Opened by fmohr 4 months ago
#68 - feat(_interfaces): modify `prepare_scoring` to allow `make_scorer` function as custom scoring
Pull Request -
State: closed - Opened by Juanpablor5 4 months ago
#67 - added LogisticRegression to search space.
Pull Request -
State: closed - Opened by fmohr 5 months ago
#66 - LogisticRegression missing among models
Issue -
State: closed - Opened by fmohr 5 months ago
#65 - Add random state to ensure reproducibility and kwargs for evaluation function
Pull Request -
State: closed - Opened by AngelG14 5 months ago
#64 - enabled possibility to exclude components or entire steps
Pull Request -
State: closed - Opened by fmohr 5 months ago
#63 - Create option to block steps and/or components from the process.
Issue -
State: closed - Opened by fmohr 5 months ago
#62 - 61 some pre processors are incompatible with pynisher and should not be executed in it
Pull Request -
State: closed - Opened by fmohr 9 months ago
#61 - Some pre-processors are incompatible with pynisher and should not be executed in it
Issue -
State: closed - Opened by fmohr 9 months ago
Labels: bug
#60 - 12 enable individual factory for custom search spaces
Pull Request -
State: closed - Opened by fmohr 9 months ago
#59 - 58 naive automl should allow proper configuration of exceptions
Pull Request -
State: closed - Opened by fmohr 10 months ago
#58 - Naive AutoML should allow proper configuration of exceptions.
Issue -
State: closed - Opened by fmohr 10 months ago
Labels: bug
#57 - Issue 46: Improved history and enabled recovery of models
Pull Request -
State: closed - Opened by fmohr 10 months ago
#56 - Add evaluation runtime and evaluation report to history
Pull Request -
State: closed - Opened by AngelG14 10 months ago
#55 - Create support for custom stoppers for early discarding
Issue -
State: open - Opened by fmohr 10 months ago
Labels: enhancement
#54 - 53 assure that svms are not executed with probabilities
Pull Request -
State: closed - Opened by fmohr 11 months ago
#53 - Assure that SVMs are not executed with probabilities.
Issue -
State: closed - Opened by fmohr 11 months ago
Labels: enhancement
#52 - enabled custom evaluators with states via update function
Pull Request -
State: closed - Opened by fmohr 11 months ago
#51 - enable stateful evaluation function
Issue -
State: closed - Opened by fmohr 11 months ago
#50 - 10 use lccv as default candidate evaluation to avoid execution timeout
Pull Request -
State: closed - Opened by AngelG14 about 1 year ago
#49 - Add evaluation runtime and evaluation report fields to history
Issue -
State: closed - Opened by fmohr about 1 year ago
Labels: enhancement
#48 - create leaderboard function
Issue -
State: closed - Opened by fmohr over 1 year ago
- 1 comment
#47 - function to recover pipeline from history
Issue -
State: closed - Opened by fmohr over 1 year ago
- 1 comment
#46 - Enrich history by additional fields
Issue -
State: closed - Opened by fmohr over 1 year ago
- 1 comment
#45 - Automatically adjust number of MCCV folds based on dataset size (variance in observations)
Issue -
State: open - Opened by fmohr over 1 year ago
Labels: enhancement
#44 - Increased NAML robustness by turning target regression object into a numeric numpy array.
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#43 - NAML is not robust with respect to uncommon target column types in regression.
Issue -
State: closed - Opened by fmohr over 1 year ago
Labels: enhancement
#42 - enable gh actions
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#41 - 38 naml infers wrong task type even though the task type could be inferred from the given scoring
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#40 - 38 naml infers wrong task type even though the task type could be inferred from the given scoring
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#39 - 38 naml infers wrong task type even though the task type could be inferred from the given scoring
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#38 - NAML infers wrong task type even though the task type could be inferred from the given scoring
Issue -
State: closed - Opened by fmohr over 1 year ago
Labels: enhancement
#37 - added case distinctions to assure that the endless loop is avoided
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#36 - added support for automated detection for the need of sparse pre-processing
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#35 - Naive AutoML does not recognize automatically whether a sparse encoding is needed
Issue -
State: closed - Opened by fmohr over 1 year ago
Labels: enhancement
#34 - exceptions can now be explicitly risen (optimizer is then stopped)
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#33 - Naive AutoML has no form of raising exceptions for debugging
Issue -
State: closed - Opened by fmohr over 1 year ago
Labels: enhancement
#32 - replaced func_timeout by pynisher
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#31 - Setup CI for improved quality
Issue -
State: closed - Opened by fmohr over 1 year ago
Labels: enhancement
#30 - Use `isinstance` to identify types of objects
Pull Request -
State: closed - Opened by PGijsbers over 1 year ago
#29 - created an check on initialization that makes sure that scorings are ok
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#28 - added __version__ to __init__.py
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#27 - added pandas to the required packages in setup.py
Pull Request -
State: closed - Opened by fmohr over 1 year ago
#26 - Naive AutoML can get stuck in infinite loop if no working pipeline is found
Issue -
State: closed - Opened by PGijsbers over 1 year ago
- 5 comments
#25 - Segmentation Fault
Issue -
State: closed - Opened by PGijsbers over 1 year ago
- 3 comments
#24 - Fix#23: Mark hpo as inactive if it stopped
Pull Request -
State: closed - Opened by PGijsbers over 1 year ago
#23 - `Cannot step inactive HPO Process` when setting `max_hpo_iterations`
Issue -
State: closed - Opened by PGijsbers over 1 year ago
#22 - Version is out of sync
Issue -
State: closed - Opened by PGijsbers over 1 year ago
#21 - Invalid scoring value results in "silent"
Issue -
State: closed - Opened by PGijsbers over 1 year ago
Labels: enhancement
#20 - Example broken with `ConfigSpace==0.7.1`
Issue -
State: closed - Opened by PGijsbers over 1 year ago
- 3 comments
Labels: bug
#19 - Package has a dependency on pandas that is not specified in `setup.py`
Issue -
State: closed - Opened by PGijsbers over 1 year ago
Labels: bug
#18 - Inactive HPO process, which should be active
Issue -
State: closed - Opened by fmohr over 1 year ago
Labels: bug
#17 - Enable three-way pipelines for transformations only on training data
Issue -
State: open - Opened by fmohr almost 2 years ago
Labels: enhancement
#16 - HPO phase is currently always strictly naive
Issue -
State: closed - Opened by fmohr almost 2 years ago
- 1 comment
Labels: bug
#15 - 8 improve history
Pull Request -
State: closed - Opened by fmohr almost 2 years ago
#14 - Add stack trace of errored evaluations to history
Issue -
State: closed - Opened by fmohr almost 2 years ago
- 1 comment
Labels: enhancement
#13 - Add Updated Default Search Space
Issue -
State: open - Opened by fmohr almost 2 years ago
Labels: enhancement
#12 - Enable individual factory for custom search spaces
Issue -
State: closed - Opened by fmohr almost 2 years ago
Labels: enhancement
#11 - Initially sort candidates by prior probability of optimal performance
Issue -
State: closed - Opened by fmohr almost 2 years ago
Labels: enhancement
#10 - Use LCCV as default candidate evaluation to avoid execution timeout
Issue -
State: closed - Opened by fmohr almost 2 years ago
Labels: enhancement
#9 - Conditional Candidate Evaluation
Issue -
State: closed - Opened by fmohr almost 2 years ago
Labels: enhancement
#8 - Improve history
Issue -
State: closed - Opened by fmohr almost 2 years ago
- 1 comment
Labels: enhancement
#7 - Enable pandas dataframes for X
Issue -
State: closed - Opened by fmohr almost 2 years ago
- 1 comment
#6 - Add option to manually specify categorical attributes
Issue -
State: closed - Opened by fmohr almost 2 years ago
- 1 comment
Labels: enhancement
#5 - Add support for class weights
Issue -
State: open - Opened by fmohr over 2 years ago
Labels: enhancement
#4 - Support for upsampling required
Issue -
State: open - Opened by fmohr over 2 years ago
Labels: enhancement
#3 - Crash when a non-preprocessed data set is used
Issue -
State: closed - Opened by lukasgehring about 3 years ago
- 1 comment
#2 - Preparing steps for pypi upload
Pull Request -
State: closed - Opened by fmohr over 3 years ago
#1 - Mlj2021
Pull Request -
State: closed - Opened by fmohr over 3 years ago