Read a horizon spreadsheet (CSV/TSV) into a PedonRecord

The everyday entry point for anyone who has a soil profile in a spreadsheet: one row per horizon, one column per attribute, using soilKey's canonical column names (top_cm, bottom_cm, designation, clay_pct, ph_h2o, munsell_hue_moist, ...). The full list is names(horizon_column_spec()); the quickest start is the bundled template system.file("extdata", "perfil_exemplo.csv", package = "soilKey").

Usage

read_pedon_csv(file, site = NULL, sep = "auto")

Arguments

file: Path to a .csv (comma) or .tsv (tab) file, one row per horizon.
site: Optional named list of site metadata (see PedonRecord). Defaults to list(id = <file base name>).
sep: Field separator. "auto" (default) uses a tab for .tsv files and a comma otherwise.

Value

A PedonRecord.

Details

Only recognised columns are used (any extras are carried through untouched), so a messy export still works: soilKey uses the columns it understands, and recognises names that differ only in capitalisation or separators and common aliases (Clay, SOC, CEC, ...). Columns and site properties that it does not recognise are named in a warning, with the closest recognised names (see unrecognised_attributes). Site metadata (id, lat/lon, soil moisture / temperature regime) is optional and passed via site; without it the profile still classifies, just with less specificity where site data would have refined the name.

Examples

f <- system.file("extdata", "perfil_exemplo.csv", package = "soilKey")
pedon <- read_pedon_csv(f)
classify_all(pedon)$summary
#>                                              wrb
#> 1 Haplic Ferralsol (Clayic, Epic, Eutric, Humic)
#>                                                                    sibcs
#> 1 Latossolos Vermelhos Distróficos típicos, textura argilosa, A moderado
#>             usda
#> 1 Typic Hapludox