Skip to contents

Generate imputation quality metric plots and ranking

Source: R/Imputation_Metrics.R
imputation_metrics.Rd

Computes NRMSE, SOR, PSS, ACC_OI via ground-truth simulation, then generates up to 6 diagnostic plots. Each plot is wrapped in tryCatch() so that a failure in one does not abort all.

Usage

imputation_metrics(
  se,
  assay_name = NULL,
  methods = .IM_BENCH_METHODS,
  combo_methods = list(),
  condition_col = "Condition",
  na_prop = 0.2,
  seed = 42L,
  pattern = "random",
  method_args = list(),
  with_value = NA_real_,
  plots = "all",
  verbose = TRUE,
  output_dir = NULL,
  export_plots = TRUE,
  export_tables = TRUE,
  plot_width = 12,
  plot_height = 8,
  plot_dpi = 150
)

Arguments

se

SummarizedExperiment (from import_imp_matrices() or pipeline).

assay_name

Character. Assay name to use as starting point. NULL (default) = first assay.

methods

Character vector of individual methods to benchmark. Default: .IM_BENCH_METHODS (16 methods). Use NULL or character(0) to skip individual methods when only using combo_methods.

combo_methods

Named list of combo/softHybrid configurations. Each element: list(mar_method, mnar_method, mode). See im_compute_metrics() for details.

condition_col

Character. Column in colData for combo methods. Default "Condition".

na_prop

Numeric (0-1). Proportion of artificial NAs. Default 0.20. Only used when pattern = "random". Ignored when pattern = "from_data" (NA proportion is derived from the data, replicating NAguideR).

seed

Integer. Random seed. Default 42.

pattern

Character. "random" or "from_data". Default "random".

method_args

Named list of per-method argument lists.

with_value

Constant for method "with".

plots

Character vector of plot names or "all" (default). Valid: "nrmse", "sor", "pss", "acc_oi", "ranking", "metrics".

verbose

Logical. Print progress. Default TRUE.

output_dir

Character. Path to export directory. If NULL (default), no files are exported. When set, tables (TSV) and/or plots (PNG) are saved to this directory (created if needed).

export_plots

Logical. Export plots as PNG when output_dir is set. Default TRUE.

export_tables

Logical. Export tables as TSV when output_dir is set. Default TRUE.

plot_width

Numeric. Width in inches for exported plots. Default 12.

plot_height

Numeric. Height in inches for exported plots. Default 8.

plot_dpi

Numeric. Resolution for exported plots. Default 150.

Value

Named list of ggplot objects (or NULL for failed plots), plus metrics_table: a data.frame from im_compute_metrics().

Examples

data(nadia_dia)
se <- im_prepare_se(nadia_dia, norm_method = "cycloess", verbose = FALSE)

# The default sweeps 16 methods; a handful is enough to illustrate it
res <- imputation_metrics(se, methods = c("min", "zero", "knn"),
                          plots = "ranking", verbose = FALSE)
res$metrics_table
#>   Method     NRMSE  SOR         PSS    ACC_OI NRMSE_Rank SOR_Rank PSS_Rank
#> 1    knn 0.1293948 1349 0.004570796 0.9916164          1        1        1
#> 2    min 3.1934374 2698 0.456162758        NA          2        2        2
#> 3   zero 6.3558709 4047 0.470781473        NA          3        3        3
#>   ACC_OI_Rank Rank_Mean
#> 1           1         1
#> 2          NA         2
#> 3          NA         3

# Two-stage combos can be compared alongside the individual methods
res2 <- imputation_metrics(se, methods = c("min", "knn"),
  combo_methods = list(
    "Impseq+min" = list(mar_method = "Impseq", mnar_method = "min")),
  plots = "ranking", verbose = FALSE)
res2$metrics_table$Method
#> [1] "knn"        "Impseq+min" "min"