Function reference • EczemaPred

Available models

Main models

BinRW: Binomial random walk model

OrderedRW: Ordered Logistic random walk model

BinMC: Binomial Markov Chain model

Reference models

MC: Markov Chain model

RW: Random walk model

Smoothing: Exponential smoothing model

AR1: Autoregressive model (order 1)

MixedAR1: Mixed effect autoregressive model (order 1)

Fitting models

EczemaModel(): EczemaModel constructor

EczemaFit(): Fit an EczemaModel

sample_prior(): Prior predictive distribution

Manipulating models

list_parameters(): List available parameters

default_prior(): Default prior

validate_prior(): Check the prior of an EczemaModel is correct

print_prior() print_distribution(): Print prior distribution

print(<EczemaModel>): Print model

prepare_standata() prepare_data_lgtd(): Prepare the data list to pass to the Stan sampler

add_prior(): Add prior to the list serving as input to the Stan sampler

replace_prior(): Replace prior

get_index() get_index2(): Associate (Patient, Time) pairs to corresponding index in the model

make_empty_data(): Create "empty" train and test set to pass to prepare_standata for sample_prior

Plotting models

plot_post_traj_pmf() plot_post_traj_fanchart() plot_ppc_traj_pmf() plot_ppc_traj_fanchart(): Plot posterior predictive trajectory

plot_latent_OrderedRW(): Plot the evolution of the expected latent score of the OrderedRW model

plot_transition_MC(): Markov Chain expected transition matrix

Validation

Utility functions for validating models

get_fc_iteration() split_fc_dataset() detail_fc_training() get_fc_training_iteration(): Forward chaining

extract_loglikelihood() extract_lpd() extract_RPS(): Extract lpd (predictive log-likelihood) and RPS from stanfit object

add_metrics1_d() add_metrics1_c() add_metrics2_d() add_metrics2_c(): Append lpd and (C)RPS to (test) dataframe

add_predictions(): Append lpd, (C)RPS and predictive samples to (test) dataframe

add_uniform_pred(): Performance of a uniform forecast

add_historical_pred(): Performance of (population) historical forecast

Miscellaneous

get_compiled_model(): Get compiled model

generate_missing(): Generate missing values in a times-series

extract_parameters(): Extract parameters posterior summary statistics

extract_simulations(): Extract simulations

samples_to_list(): Process samples to a list that can be included to a dataframe

add_fanchart(): Add fanchart to ggplot

add_broken_pointline(): Add broken pointline to ggplot

EczemaPred-package EczemaPred: The 'EczemaPred' package.