Get columns shared by most datasets — get_meta

In most trialmaster exports, many datasets share a certain amount of columns containing meta-data that are often irrelevant to the point. This function identifies the columns that are present in at least 95% of datasets (by default)

Usage

get_meta_cols(min_pct = getOption("edc_meta_cols_pct", 0.95))

Arguments

min_pct: Default=0.95. The minimal proportion of datasets a column has to reach. Subject ID is always excluded.

Value

a character vector

Examples

tm = edc_example()
#> Warning: Option "edc_lookup" has been overwritten.
load_list(tm)
meta_cols = get_meta_cols()
long_mixed %>% dplyr::select(-dplyr::any_of(meta_cols))
#> # A tibble: 100 × 4
#>    subjid    val1b val2b val3b
#>     <int>    <dbl> <dbl> <chr>
#>  1      1  1.33    11.0  B    
#>  2      1 -0.869   10.9  B    
#>  3      2  0.0555  10.0  C    
#>  4      2  0.0491  10.1  C    
#>  5      3 -0.578    9.28 D    
#>  6      3 -0.999    9.80 D    
#>  7      4 -0.00243  8.97 E    
#>  8      4  0.656    9.03 E    
#>  9      5  1.48     8.78 F    
#> 10      5 -1.91    10.8  F    
#> # ℹ 90 more rows