get_pEti_samp_reg: get etiology samples by names (no regression)
In oslerinhealth/baker: Bayesian Analysis Kit for Etiology Research

View source: R/plot-etiology-side-by-side.R

get etiology samples by names (no regression)

get_pEti_samp_reg(
  DIR_NPLCM,
  model_options,
  stratum_bool = NULL,
  pEti_subject = FALSE,
  reg_param = FALSE,
  truth = NULL,
  return_metric = FALSE,
  RES_NPLCM = NULL
)

`DIR_NPLCM`	File path to the folder containing posterior samples
`model_options`	See `nplcm()`
`stratum_bool`	a vector of TRUE/FALSE with TRUE indicating the rows of subjects to include
`pEti_subject`	TRUE for getting individual specific samples; Default to `FALSE`
`reg_param`	TRUE for getting regression parameters; Default to `FALSE`
`truth`	a list of truths computed from true parameters in simulations; elements: Eti, FPR, PR_case,TPR; All default to `NULL` in real data analyses. Currently only works for one slice of bronze-standard measurements (in a non-nested model). Eti matrix of # of rows = # of subjects, # columns: `length(cause_list)` for Eti FPR matrix of # of rows = # of subjects, # columns: `ncol(data_nplcm$Mobs$MBS$MBS1)` PR_case matrix of # of rows = # of subjects, # columns: `ncol(data_nplcm$Mobs$MBS$MBS1)` TPR a vector of length identical to `PR_case`
`return_metric`	TRUE for showing overall mean etiology, quantiles, s.d., and if `truth$Eti` is supplied, coverage, bias, truth and integrated mean squared errors (IMSE).
`RES_NPLCM`	pre-read res_nplcm; default to NULL to save time.

A list:

pEti_mat: a matrix of posterior samples (iteration by cause); overall etiology latent_nm: a vector of character strings representing the names of the causes pEti_subject: an array of individual specific etiology; cause by subject by iteration betaEti: an array of samples of the etiology regression parameters; coefficient by cause by iteration

oslerinhealth/baker documentation built on May 22, 2021, 12:05 p.m.