BayesMallows (version 1.5.0)

compute_consensus.SMCMallows: Compute Consensus Ranking

Description

Compute the consensus ranking using either cumulative probability (CP) or maximum a posteriori (MAP) consensus vitelli2018BayesMallows. For mixture models, the consensus is given for each mixture.

Usage

# S3 method for SMCMallows
compute_consensus(model_fit, type = "CP", ...)

Arguments

model_fit

An object of class SMCMallows, returned from smc_mallows_new_item_rank or smc_mallows_new_users.

type

Character string specifying which consensus to compute. Either "CP" or "MAP". Defaults to "CP".

...

Other optional arguments passed to methods. Currently not used.

See Also

Other posterior quantities: assign_cluster(), compute_consensus.BayesMallows(), compute_consensus(), compute_posterior_intervals.BayesMallows(), compute_posterior_intervals.SMCMallows(), compute_posterior_intervals(), heat_plot(), plot.BayesMallows(), plot.SMCMallows(), plot_elbow(), plot_top_k(), predict_top_k(), print.BayesMallowsMixtures(), print.BayesMallows()

Examples

# Basic elements
data <- sushi_rankings[1:100, ]
n_items <- ncol(data)
leap_size <- floor(n_items / 5)
metric <- "footrule"
Time <- 20
N <- 100

# Prepare exact partition function
cardinalities <- prepare_partition_function(metric = metric,
                                            n_items = n_items)$cardinalities

# Performing SMC
smc_test <- smc_mallows_new_users(
  R_obs = data, type = "complete", n_items = n_items,
  metric = metric, leap_size = leap_size,
  N = N, Time = Time,
  cardinalities = cardinalities,
  mcmc_kernel_app = 5,
  num_new_obs = 5,
  alpha_prop_sd = 0.5,
  lambda = 0.15,
  alpha_max = 1e6
)

compute_posterior_intervals(smc_test, parameter = "rho")

compute_consensus(model_fit = smc_test, type = "CP")
compute_consensus(model_fit = smc_test, type = "MAP")

compute_posterior_intervals(smc_test, parameter = "alpha")