Tuning Parameters | Parameter Objects | Numeric Parameters | Discrete Parameters | Creating Novel Parameters | Unknown Values | Parameter Sets | Parameter Grids
Package: dials 1.4.4.9000
Hannah Frick
dials: Tools for Creating Tuning Parameter Values
Many models contain tuning parameters (i.e. parameters that cannot be directly estimated from the data). These tools can be used to define objects for creating, simulating, or validating values for such parameters.
Authors:
dials_1.4.4.9000.tar.gz
dials_1.4.4.9000.zip(r-4.7)dials_1.4.4.9000.zip(r-4.6)dials_1.4.4.9000.zip(r-4.5)
dials_1.4.4.9000.tgz(r-4.6-any)dials_1.4.4.9000.tgz(r-4.5-any)
dials_1.4.4.9000.tar.gz(r-4.7-any)dials_1.4.4.9000.tar.gz(r-4.6-any)
dials_1.4.4.9000.tgz(r-4.6-emscripten)
manual.pdf |manual.html✨
DESCRIPTION |NEWS
card.svg |card.png
dials/json (API)
| # Install 'dials' in R: |
| install.packages('dials', repos = c('https://tidymodels.r-universe.dev', 'https://cloud.r-project.org')) |
Bug tracker:https://github.com/tidymodels/dials/issues
Pkgdown/docs site:https://dials.tidymodels.org
Last updated from:0c88e88546. Checks:9 OK. Indexed: yes.
| Target | Result | Time | Files | Syslog |
|---|---|---|---|---|
| linux-devel-x86_64 | OK | 196 | ||
| source / vignettes | OK | 192 | ||
| linux-release-x86_64 | OK | 157 | ||
| macos-release-arm64 | OK | 114 | ||
| macos-oldrel-arm64 | OK | 143 | ||
| windows-devel | OK | 158 | ||
| windows-release | OK | 155 | ||
| windows-oldrel | OK | 182 | ||
| wasm-release | OK | 107 |
Exports:activationactivation_2adjust_deg_freeall_neighborsattention_typeaverage_before_softmaxbalance_probabilitiesbatch_sizebottleneck_unitsbuffercal_method_classcal_method_regclass_weightsconditional_min_criterionconditional_test_statisticconditional_test_typeconfidence_factorcostcost_complexitydeg_freedegreedegree_intdiagonal_covariancedist_powerdropoutdropout_attndropout_embeddingdropout_hiddendropout_lastencode_unitepochsextract_parameter_dialsextrapolationfinalizefreq_cutfuzzy_thresholdingget_batch_sizesget_log_pget_nget_n_fracget_n_frac_rangeget_pget_rbf_rangegrid_latin_hypercubegrid_max_entropygrid_randomgrid_regulargrid_space_fillingharmonic_frequencyhas_unknownshidden_unitshidden_units_2initial_umapis_unknownkernel_offsetl2_leaf_regLaplacelearn_rateloss_reductionlower_limitlower_quantilemax_leavesmax_nodesmax_num_termsmax_rulesmax_timesmax_tokensmin_distmin_nmin_timesmin_uniquemixturemomentummtrymtry_longmtry_propneighborsnew_qual_paramnew_quant_paramno_global_pruningnormalizationnum_attn_blocksnum_attn_featnum_attn_headsnum_breaksnum_clustersnum_compnum_embeddingnum_estimatorsnum_hashnum_knotsnum_leavesnum_random_splitsnum_runsnum_termsnum_tokensodds_linkordinal_linkover_ratioparametersparameters_constrpenaltypenalty_averagepenalty_L1penalty_L2penalty_typepredictor_proppredictor_winnowingprior_mixture_thresholdprior_outcome_rangeprior_slab_dispersionprior_terminal_node_coefprior_terminal_node_expoprod_degreeprop_termspruneprune_methodrange_getrange_setrange_validateranger_class_rulesranger_reg_rulesranger_split_rulesranger_survival_rulesrate_decayrate_initialrate_largestrate_reductionrate_schedulerate_step_sizerate_stepsrbf_sigmaregularization_factorregularization_methodregularize_depthresid_atrule_bandssample_propsample_sizescale_factorscale_pos_weightselect_featuresshrinkage_correlationshrinkage_frequenciesshrinkage_variancesigned_hashsignificance_thresholdsmoothnesssoftmax_temperaturespline_degreesplitting_rulestep_ratestop_itersummary_statsurv_distsurvival_linksvm_margintarget_tokentarget_weightthresholdtokentraining_set_limittree_depthtreestrim_amountunbiased_rulesunder_ratiounique_cutunknownupper_limitvalidation_set_propvalue_inversevalue_samplevalue_seqvalue_setvalue_transformvalue_validatevalues_activationvalues_attention_typevalues_cal_clsvalues_cal_regvalues_initial_umapvalues_normalizationvalues_odds_linkvalues_ordinal_linkvalues_penalty_typevalues_prune_methodvalues_regularization_methodvalues_schedulervalues_summary_statvalues_surv_distvalues_survival_linkvalues_test_statisticvalues_test_typevalues_tokenvalues_weight_funcvalues_weight_schemevocabulary_sizeweightweight_funcweight_schemewindow_size
Dependencies:cliDiceDesigndplyrfarvergenericsgluehardhatlabelinglifecyclemagrittrpillarpkgconfigpurrrR6RColorBrewerrlangscalessfdsparsevctrstibbletidyselectutf8vctrsviridisLitewithr
Readme and manuals
Help Manual
| Help page | Topics |
|---|---|
| Activation functions between network layers | activation activation_2 values_activation |
| Parameters to adjust effective degrees of freedom | adjust_deg_free |
| Parameter to determine which neighbors to use | all_neighbors |
| Parameters for attention-based tabular models | attention-param attention_type dropout_attn normalization num_attn_blocks num_attn_feat num_attn_heads target_token values_attention_type values_normalization |
| Parameters for BART models These parameters are used for constructing Bayesian adaptive regression tree (BART) models. | bart-param prior_outcome_range prior_terminal_node_coef prior_terminal_node_expo |
| Buffer size | buffer |
| Methods for model calibration | cal_method_class cal_method_reg values_cal_cls values_cal_reg |
| Parameters for class weights for imbalanced problems | class_weights |
| Parameters for possible engine parameters for partykit models | conditional_min_criterion conditional_test_statistic conditional_test_type values_test_statistic values_test_type |
| Parameters for possible engine parameters for C5.0 | confidence_factor fuzzy_thresholding no_global_pruning predictor_winnowing rule_bands |
| Support vector machine parameters | cost svm_margin |
| Degrees of freedom (integer) | deg_free |
| Parameters for exponents | degree degree_int prod_degree spline_degree |
| Minkowski distance parameter | dist_power |
| Neural network parameters | batch_size dropout dropout_embedding dropout_hidden dropout_last epochs hidden_units hidden_units_2 num_embedding |
| Parameters for possible engine parameters for Cubist | extrapolation max_rules unbiased_rules |
| Functions to finalize data-specific parameter ranges | finalize finalize.default finalize.list finalize.logical finalize.param finalize.parameters get_log_p get_n get_n_frac get_n_frac_range get_p get_rbf_range |
| Near-zero variance parameters | freq_cut unique_cut |
| Create grids of tuning parameters | grid_random grid_random.default grid_random.list grid_random.param grid_random.parameters grid_regular grid_regular.default grid_regular.list grid_regular.param grid_regular.parameters |
| Space-filling parameter grids | grid_space_filling grid_space_filling.default grid_space_filling.list grid_space_filling.param grid_space_filling.parameters |
| Harmonic Frequency | harmonic_frequency |
| Initialization method for UMAP | initial_umap values_initial_umap |
| Possible engine parameters for catboost | l2_leaf_reg max_leaves |
| Laplace correction parameter | Laplace |
| Learning rate | learn_rate |
| Parameters for possible engine parameters for randomForest | max_nodes |
| Parameters for possible engine parameters for earth models | max_num_terms |
| Word frequencies for removal | max_times min_times |
| Maximum number of retained tokens | max_tokens |
| Parameter for the effective minimum distance between embedded points | min_dist |
| Number of unique values for pre-processing | min_unique |
| Mixture of penalization terms | mixture |
| Gradient descent momentum parameter | momentum |
| Number of randomly sampled predictors | mtry mtry_long |
| Proportion of Randomly Selected Predictors | mtry_prop |
| Number of neighbors | neighbors |
| Tools for creating new parameter objects | new-param new_qual_param new_quant_param |
| Number of cut-points for binning | num_breaks |
| Number of Clusters | num_clusters |
| Number of new features | num_comp num_terms |
| Text hashing parameters | num_hash signed_hash |
| Number of knots (integer) | num_knots |
| Possible engine parameters for lightbgm | num_leaves |
| Number of Computation Runs | num_runs |
| Parameter to determine number of tokens in ngram | num_tokens |
| Ordinal Regression Link Functions (character) | odds_link ordinal_link values_odds_link values_ordinal_link |
| Parameters for class-imbalance sampling | over_ratio under_ratio |
| Create a parameter set | parameters parameters.default parameters.list parameters.param |
| Amount of regularization/penalization | penalty |
| Proportion of predictors | predictor_prop |
| Bayesian PCA parameters | prior_mixture_threshold prior_slab_dispersion |
| Proportion of top predictors | prop_terms |
| MARS pruning methods | prune_method values_prune_method |
| Limits for the range of predictions | lower_limit range_limits upper_limit |
| Tools for working with parameter ranges | range_get range_set range_validate |
| Kernel parameters | kernel_offset rbf_sigma scale_factor |
| Parameters for possible engine parameters for ranger | lower_quantile num_random_splits ranger_class_rules ranger_reg_rules ranger_split_rules ranger_survival_rules regularization_factor regularize_depth significance_threshold splitting_rule |
| Estimation methods for regularized models | regularization_method values_regularization_method |
| Parameters for residual networks | bottleneck_units resid_at resnet-param |
| Parameters for regularization learning networks | penalty_average penalty_type rln-param step_rate values_penalty_type |
| Parameters for possible engine parameters for xgboost | penalty_L1 penalty_L2 scale_pos_weight |
| Parameters for neural network learning rate schedulers | rate_decay rate_initial rate_largest rate_reduction rate_schedule rate_steps rate_step_size scheduler-param values_scheduler |
| Parameter to enable feature selection | select_features |
| Parameters for possible engine parameters for sda models | diagonal_covariance shrinkage_correlation shrinkage_frequencies shrinkage_variance |
| Kernel Smoothness | smoothness |
| Early stopping parameter | stop_iter |
| Rolling summary statistic for moving windows | summary_stat values_summary_stat |
| Parametric distributions for censored data | surv_dist values_surv_dist |
| Survival Model Link Function | survival_link values_survival_link |
| Parameters for TabPFN models | average_before_softmax balance_probabilities num_estimators softmax_temperature tab-pfn-param training_set_limit |
| Amount of supervision parameter | target_weight |
| General thresholding parameter | threshold |
| Token types | token values_token |
| Parameter functions related to tree- and rule-based models. | cost_complexity loss_reduction min_n prune sample_prop sample_size trees tree_depth |
| Amount of Trimming | trim_amount |
| Placeholder for unknown parameter values | has_unknowns is_unknown unknown |
| Update a single parameter in a parameter set | update.parameters |
| Proportion of data used for validation | validation_set_prop |
| Tools for working with parameter values | value_inverse value_sample value_seq value_set value_transform value_validate |
| Number of tokens in vocabulary | vocabulary_size |
| Parameter for '"double normalization"' when creating token counts | weight |
| Kernel functions for distance weighting | values_weight_func weight_func |
| Term frequency weighting methods | values_weight_scheme weight_scheme |
| Parameter for the moving window size | window_size |
