The tidy() method for RuleFit models was not using the penalty value. This is corrected and a single penalty value is required for using the function. (#66)
Fixed bug where predict sometimes didn't work for xrf models.
C50 tree tidy() method nor works with threes by showing terminal nodes in rules.
tidy() method for Cubist models now has an option for how many committees to tidy.
Added a tidy() method for C5.0 rule-based models.
The mtry_prop parameter was moved to the dials package and is now re-exported here for backward compatibility.
A bug was fixed related to multi_predict() with C5.0 rule-based models (#49).
The mtry argument is now mapped to colsample_bynode rather than colsample_bytree. This is consistent with parsnip's interface to xgboost as of parsnip 0.1.6. colsample_bytree can still be optimized by passing it in as an engine argument to set_engine() (#60).
Introduced support for early stopping in rule_fit() via the stop_iter argument. See parsnip::details_rule_fit_xrf. Note that this is a main argument to rule_fit() requiring parsnip 1.0.0.
cubist_rules()) were moved to the parsnip package.Added tidy() methods for cubist and RuleFit models to get back the rules and model coefficients.
Clean up tunable() methods that define the default parameter ranges for the tuning parameters.
Changes to test for cases when CRAN cannot get xgboost to work on their Solaris configuration.
Case weights were enabled for [C5_rules()] and [cubist_rules()].
parsnip will respect the encoding that the underlying model uses.Documentation changes (some demanded by CRAN).
Changes to make rules work well with PSOCK cluster parallelism.
dplyr version.