parsnip 0.0.5

Fixes

• A bug (#206 and #234) was fixed that caused an error when predicting with a multinomial glmnet model.

Other Changes

• glmnet was removed as a dependency since the new version depends on 3.6.0 or greater. Keeping it would constrain parsnip to that same requirement. All glmnet tests are run locally.

• A set of internal functions are now exported. These are helpful when creating a new package that registers new model specifications.

New Features

• nnet was added as an engine to multinom_reg() #209

Breaking Changes

• There were some mis-mapped parameters (going between parsnip and the underlying model function) for spark boosted trees and some keras models. See 897c927.

parsnip 0.1.2

Breaking Changes

• parsnip now has options to set specific types of predictor encodings for different models. For example, ranger models run using parsnip and workflows do the same thing by not creating indicator variables. These encodings can be overridden using the blueprint options in workflows. As a consequence, it is possible to get a different model fit that previous versions of parsnip. More details about specific encoding changes are below. (#326)

Other Changes

• tidyr >= 1.0.0 is now required.

• SVM models produced by kernlab now use the formula method (see breaking change notice above). This change was due to how ksvm() made indicator variables for factor predictors (with one-hot encodings). Since the ordinary formula method did not do this, the data are passed as-is to ksvm() so that the results are closer to what one would get if ksmv() were called directly.

• MARS models produced by earth now use the formula method.

• For xgboost, a one-hot encoding is used when indicator variables are created.

• Under-the-hood changes were made so that non-standard data arguments in the modeling packages can be accommodated. (#315)

New Features

• A new main argument was added to boost_tree() called stop_iter for early stopping. The xgb_train() function gained arguments for early stopping and a percentage of data to leave out for a validation set.

• If fit() is used and the underlying model uses a formula, the actual formula is pass to the model (instead of a placeholder). This makes the model call better.

• A function named repair_call() was added. This can help change the underlying models call object to better reflect what they would have obtained if the model function had been used directly (instead of via parsnip). This is only useful when the user chooses a formula interface and the model uses a formula interface. It will also be of limited use when a recipes is used to construct the feature set in workflows or tune.

• The predict() function now checks to see if required modeling packages are installed. The packages are loaded (but not attached). (#249) (#308) (tidymodels/workflows#45)

• The function req_pkgs() is a user interface to determining the required packages. (#308)

parsnip 0.1.1

New Features

• liquidSVM was added as an engine for svm_rbf() (#300)

Fixes

• The error message for missing packages was fixed (#289 and #292)

Other Changes

• S3 dispatch for tidy() was broken on R 4.0.

parsnip 0.1.0

• Release for enabling more parsnip-adjacent packages.

• Stan models no longer generate output when fit. Use the refresh option when setting the engine to change this.

• Data sets have been moved to the modeldata package.

CRAN release v0.0.4

Merge pull request #230 from tidymodels/control-parsnip

new control function name

CRAN release v0.0.3.1

No long double changes for CRAN submission (#203)

* better handling of unknown modes

* version sent to cran

CRAN release v0.0.3

cran-release-0.0.3

added more binaries

CRAN release v0.0.1

cran-release-0.0.1

remove duplicate lines

CRAN release candidate for v0.0.1

cran-candidate-0.0.1

fixes for r-hub checks

snapshot prior to removal of recipes code

parsnip will be agnostic to recipes after this release of the master branch.

The reason is that the recipe would have to be bound to the fitted model.

That's a bad idea because future code should be able to train a recipe and then use it across many different models (or sub-models). The version is this release is tagged because the existing code can be recycled into the other packages that will need to use recipes for different models.