Skip to contents

Parse messy numeric result strings into a structured tibble

Source: R/numeric_parser.R
parse_numeric_results.Rd

Handles whitespace, commas, scientific notation (x10 with or without caret and spaces, spaced E, Fortran exponents, standard e), word multipliers ("7 million"), footnote asterisks and doubled-decimal typos, qualifier extraction (<, >, <=, >=, ~, =), unicode qualifiers (>=, <=), narrative detection (BDL, ND, trace, etc.), range splitting (5-10 -> 3 rows), and unparseable values.

Usage

parse_numeric_results(values)

Arguments

values

Character vector of raw result strings

Value

A tibble with columns: orig_row_id, orig_result, numeric_value, qualifier, range_bin, parse_flag. Range values produce 3 rows sharing the same orig_row_id.

Details

Unparseable strings containing no digits (e.g. "See note") are reclassified as narrative: numeric_value = NA, parse_flag = "narrative", excluded from the unparseable warning and the correction workflow.

Range splitting (PARS-03): "5-10" becomes 3 rows per D-02 and D-03:

  • low row: qualifier=">=", range_bin="low"

  • mid row: qualifier="~", range_bin="mid"

  • high row: qualifier="<=", range_bin="high" Negative numbers (-5) and scientific notation (1e-3) are NOT split (numeric pre-guard).

Implementation note: range detection runs on a pre-Fortran-normalized form. This is critical because the Fortran exponent normalizer converts "5-10" to "5e-10" (matches digit-minus-digits pattern). Ranges are detected first, then Fortran normalization applies only to non-range values.

Examples

parse_numeric_results(c("< 5.0", "2.5x10^3", "BDL", "4.56+02", "\u2265100"))
#> # A tibble: 5 × 6
#>   orig_row_id orig_result numeric_value qualifier range_bin parse_flag 
#>         <int> <chr>               <dbl> <chr>     <chr>     <chr>      
#> 1           1 < 5.0                   5 "<"       as_is     ""         
#> 2           2 2.5x10^3             2500 ""        as_is     ""         
#> 3           3 BDL                    NA ""        as_is     "narrative"
#> 4           4 4.56+02               456 ""        as_is     ""         
#> 5           5 ≥100                  100 ">="      as_is     ""         
parse_numeric_results(c("5-10", "-5", "1e-3"))
#> # A tibble: 5 × 6
#>   orig_row_id orig_result numeric_value qualifier range_bin parse_flag
#>         <int> <chr>               <dbl> <chr>     <chr>     <chr>     
#> 1           1 5-10                5     ">="      low       ""        
#> 2           1 5-10                7.5   "~"       mid       ""        
#> 3           1 5-10               10     "<="      high      ""        
#> 4           2 -5                 -5     ""        as_is     ""        
#> 5           3 1e-3                0.001 ""        as_is     ""