Skip to contents

Coerce IRS Binary-Indicator Columns to Logical

Source: R/nccs_as_indicator.R
nccs_as_indicator.Rd

IRS Business Master File extracts encode binary fields using inconsistent tokens that vary by vintage. This helper coerces such a column to a logical vector under one of two well-known schemes.

Usage

nccs_as_indicator(x, scheme = c("yn", "efile"))

Arguments

x

A character, numeric, or logical vector.

scheme

One of `"yn"` (default) or `"efile"`.

Value

A logical vector the same length as `x`.

Schemes

`"yn"` — general yes/no indicator columns (often the `_cd` suffix in upstream data).

  • TRUE: `Y`, `y`, `1`, `T`, `TRUE`, `true`, `True`

  • FALSE: `N`, `n`, `0`, `2`, `F`, `FALSE`, `false`, `False`

Note that `"2"` maps to FALSE: the IRS shifted some binary fields to a 1 (= yes) / 2 (= no) encoding in recent vintages (e.g. some 2022/2023 990 columns).

`"efile"` — the IRS e-file indicator, whose accepted tokens vary by filing year:

  • 2015 990 / 990-EZ used `E` (electronic) / `P` (paper).

  • 2016 and 2017 990 / 990-EZ switched to `Y` / `N`.

  • 2018 onwards returned to `E` / `P`.

All three encodings are accepted:

  • TRUE: `E`, `e`, `Y`, `y`, `1`, `T`, `TRUE`, `true`, `True`

  • FALSE: `P`, `p`, `N`, `n`, `0`, `F`, `FALSE`, `false`, `False`

Tokens outside the accepted set become `NA`, with a single `warning()` that lists the distinct unknown values (not one per element). `NA` input is propagated silently.

Pure base-R, no side effects, no in-place mutation. Designed for use on external data the user is joining against [nccs_read()] output — columns returned by `nccs_read()` are already cleaned upstream.

Examples

nccs_as_indicator(c("Y", "N", "1", "2", NA))
#> [1]  TRUE FALSE  TRUE FALSE    NA
# [1]  TRUE FALSE  TRUE FALSE    NA

nccs_as_indicator(c("E", "P", "Y", "N"), scheme = "efile")
#> [1]  TRUE FALSE  TRUE FALSE
# [1]  TRUE FALSE  TRUE FALSE