Read the life-history timing table, filtered by group and species

One row per watershed group, species, life stage and window. start and end are month-day (MM-DD) with no year, so a window with start > end crosses 1 January. A window the source gave only one end of keeps the other as NA; a row with neither is an error. source holds Zotero citation keys separated by "; ".

Usage

knw_timing_read(path, wsg_code = NULL, species_code = NULL)

Arguments

path: character(1) path to the timing CSV, with columns wsg_code, species_code, life_stage, start, end, source and note.
wsg_code: character watershed group codes to keep, or NULL for all.
species_code: character species codes to keep, or NULL for all.

Value

A tibble with the seven columns, all character.

Details

The table is validated before filtering, so a bad row anywhere in the file is an error whichever group is asked for. Within one group, species and stage, windows must not overlap (a half window's known day counts), and a window crossing 1 January is one row: a window ending 12-31 beside one starting 01-01 is refused, because the pair reads as two windows to anything that counts them. start == end is refused as ambiguous between one day and a whole year. A requested group or species absent from the table is a warning, so a typo does not read as "no data".

Examples

path <- tempfile(fileext = ".csv")
utils::write.csv(data.frame(
  wsg_code = c("MORR", "MORR"), species_code = c("CH", "BT"),
  life_stage = c("spawning", "incubation"), start = c("09-01", "08-23"),
  end = c("10-07", "04-30"), source = c("key_a", "key_b"), note = NA
), path, row.names = FALSE, na = "")
knw_timing_read(path, wsg_code = "MORR", species_code = "BT")
#> # A tibble: 1 × 7
#>   wsg_code species_code life_stage start end   source note 
#>   <chr>    <chr>        <chr>      <chr> <chr> <chr>  <chr>
#> 1 MORR     BT           incubation 08-23 04-30 key_b  NA