R/sample-info.R

In facileverse/FacileData: A fluent API for accessing multi-assay high-throughput genomics data

#' Fetch the sample statistics for sets of samples in the warehouse
#'
#' NOTE: this function needs the axe. It has been changed to use the
#' assay_sample_info_table, but the way we handle this with the new unhinged
#' assay needs to change.
#'
#' @export
#' @param x A \code{FacileDataSet} object
#' @param samples a data.frame or tbl_sqlite that has dataset and sample_id
#'   columns
#' @param semi use \code{semi_join}? I've found this to be slow sometimes in
#'   SQLite for some reason
#' @param assay_name parameter added to keep old API same with new "unhinged"
#'   FacileDataSets.
#' @return a tbl_df or tbl_sqlite result from the sample_stats table
#' @family API
fetch_sample_statistics.FacileDataSet <- function(x, samples = NULL,
                                                  semi = TRUE,
                                                  assay_name = default_assay(x),
                                                  ...) {
  warning("`fetch_sample_statistics` will be removed from FacileData API\n,",
          "See Issue #2\n",
          "https://github.com/denalitherapeutics/FacileData/issues/2",
          immediate. = TRUE)

  assert_string(assay_name)
  stopifnot(assay_name %in% assay_names(x))

  ss <- assay_sample_info_tbl(x) |>
    filter(assay == assay_name) |>
    set_fds(x)

  if (is.null(samples)) {
    out <- ss
  } else {
    # TODO: Need to write unit tests here to exercise what we want to do with
    #       these results when samples are provided
    samples <- assert_sample_subset(samples)
    out <- join_samples(ss, samples, semi)
  }

  set_fds(out, x)
}