Function reference • dials

Parameter Sets

parameters(): Information on tuning parameters within an object

update(<parameters>): Update a single parameter in a parameter set

range_validate() range_get() range_set(): Tools for working with parameter ranges

value_validate() value_seq() value_sample() value_transform() value_inverse() value_set(): Tools for working with parameter values

Grid Creation

grid_max_entropy() grid_latin_hypercube(): Space-filling parameter grids

grid_regular() grid_random(): Create grids of tuning parameters

Parameter Objects

activation() values_activation: Activation functions between network layers

adjust_deg_free(): Parameters to adjust effective degrees of freedom

all_neighbors(): Parameter to determine which neighbors to use

prior_terminal_node_coef() prior_terminal_node_expo() prior_outcome_range(): Parameters for BART models These parameters are used for constructing Bayesian adaptive regression tree (BART) models.

class_weights(): Parameters for class weights for imbalanced problems

conditional_min_criterion() values_test_type conditional_test_type() values_test_statistic conditional_test_statistic(): Parameters for possible engine parameters for partykit models

confidence_factor() no_global_pruning() predictor_winnowing() fuzzy_thresholding() rule_bands(): Parameters for possible engine parameters for C5.0

cost() svm_margin(): Support vector machine parameters

deg_free(): Degrees of freedom (integer)

degree() degree_int() spline_degree() prod_degree(): Parameters for exponents

dist_power(): Minkowski distance parameter

dropout() epochs() hidden_units() batch_size(): Neural network parameters

extrapolation() unbiased_rules() max_rules(): Parameters for possible engine parameters for Cubist

freq_cut() unique_cut(): Near-zero variance parameters

harmonic_frequency(): Harmonic Frequency

Laplace(): Laplace correction parameter

learn_rate(): Learning rate

max_nodes(): Parameters for possible engine parameters for randomForest

max_num_terms(): Parameters for possible engine parameters for earth models

max_times() min_times(): Word frequencies for removal

max_tokens(): Maximum number of retained tokens

min_dist(): Parameter for the effective minimum distance between embedded points

min_unique(): Number of unique values for pre-processing

mixture(): Mixture of penalization terms

momentum(): Gradient descent momentum parameter

mtry() mtry_long(): Number of randomly sampled predictors

mtry_prop(): Proportion of Randomly Selected Predictors

neighbors(): Number of neighbors

num_breaks(): Number of cut-points for binning

num_clusters(): Number of Clusters

num_comp() num_terms(): Number of new features

num_hash() signed_hash(): Text hashing parameters

num_knots(): Number of knots (integer)

num_leaves(): Possible engine parameters for lightbgm

num_runs(): Number of Computation Runs

num_tokens(): Parameter to determine number of tokens in ngram

over_ratio() under_ratio(): Parameters for class-imbalance sampling

penalty(): Amount of regularization/penalization

predictor_prop(): Proportion of predictors

prior_slab_dispersion() prior_mixture_threshold(): Bayesian PCA parameters

prune_method() values_prune_method: MARS pruning methods

rate_initial() rate_largest() rate_reduction() rate_steps() rate_step_size() rate_decay() rate_schedule() values_scheduler: Parameters for neural network learning rate schedulers These parameters are used for constructing neural network models.

rbf_sigma() scale_factor() kernel_offset(): Kernel parameters

regularization_factor() regularize_depth() significance_threshold() lower_quantile() splitting_rule() ranger_class_rules ranger_reg_rules ranger_split_rules num_random_splits(): Parameters for possible engine parameters for ranger

regularization_method() values_regularization_method: Estimation methods for regularized models

scale_pos_weight() penalty_L2() penalty_L1(): Parameters for possible engine parameters for xgboost

select_features(): Parameter to enable feature selection

shrinkage_correlation() shrinkage_variance() shrinkage_frequencies() diagonal_covariance(): Parameters for possible engine parameters for sda models

smoothness(): Kernel Smoothness

stop_iter(): Early stopping parameter

summary_stat() values_summary_stat: Rolling summary statistic for moving windows

surv_dist() values_surv_dist: Parametric distributions for censored data

survival_link() values_survival_link: Survival Model Link Function

threshold(): General thresholding parameter

token() values_token: Token types

trees() min_n() sample_size() sample_prop() loss_reduction() tree_depth() prune() cost_complexity(): Parameter functions related to tree- and rule-based models.

trim_amount(): Amount of Trimming

validation_set_prop(): Proportion of data used for validation

vocabulary_size(): Number of tokens in vocabulary

weight(): Parameter for "double normalization" when creating token counts

weight_func() values_weight_func: Kernel functions for distance weighting

weight_scheme() values_weight_scheme: Term frequency weighting methods

window_size(): Parameter for the moving window size

Finalizing Parameters

finalize() get_p() get_log_p() get_n_frac() get_n_frac_range() get_n() get_rbf_range() get_batch_sizes(): Functions to finalize data-specific parameter ranges

Developer Tools

encode_unit(): Class for converting parameter values back and forth to the unit range

new_quant_param() new_qual_param(): Tools for creating new parameter objects

parameters_constr(): Construct a new parameter set object

unknown() is_unknown() has_unknowns(): Placeholder for unknown parameter values