Skip to contents

Subset an AlphaVarSet Object

Source: R/10_dataops.R
av_subset.Rd

Filter the data in an AlphaVarSet while tracking filter history and clearing stale statistics. This enables dynamic drill-down analysis.

Usage

av_subset(obj, ..., .description = NULL)

Arguments

obj

An AlphaVarSet object.

...

Filter expressions passed to dplyr::filter().

.description

Optional character string describing this filter step (for breadcrumb display).

Value

A new AlphaVarSet object with:

  • Filtered data

  • Cleared stats slot (to prevent stale statistics)

  • Updated meta$filter_history with the filter breadcrumb

Details

This function is the core "scalpel" for the dynamic exploration engine. Each call creates a new subset and records the filter in the history, allowing users to trace their analysis path.

Statistics are intentionally cleared because global stats computed on the full dataset would be misleading when applied to a subset. Use av_calc_enrichment() after subsetting to compute fresh statistics.

Examples

example_file <- system.file("extdata", "bcl11a_rs1427407_example.csv",
                            package = "AlphaVaR")
obj <- av_import_csv(example_file)
#> Reading bcl11a_rs1427407_example.csv ...

atac <- av_subset(obj, output_type == "ATAC", .description = "ATAC tracks")
nrow(atac$data)
#> [1] 320

# The score column is metadata, so inject it instead of naming it literally.
sc <- obj$meta$active_score
strong_atac <- av_subset(atac, !!rlang::sym(sc) > 0.9,
                         .description = "quantile > 0.9")
nrow(strong_atac$data)
#> [1] 97

av_get_filter_history(strong_atac)
#> # A tibble: 2 × 5
#>    step description    n_before n_after timestamp          
#>   <dbl> <chr>             <int>   <int> <dttm>             
#> 1     1 ATAC tracks        2676     320 2026-08-14 20:02:35
#> 2     2 quantile > 0.9      320      97 2026-08-14 20:02:35