Convert a datras_raw object to a wide-format table

Create a wide-format data frame from a datras_raw / DATRASraw object, using HH as the main haul-level table and spreading selected HL variables into species-specific columns.

Usage

as_wide_format(
  x,
  vars_hh = c("Survey", "Gear", "Country", "Ship", "Year", "Quarter", "Month", "Day",
    "lon", "lat", "timeOfYear", "abstime", "DayNight", "TimeShotHour", "HaulDur"),
  vars_hl = "Count",
  species_var = "Species",
  id_var = "haul.id",
  sep = "__",
  fill = 0,
  sanitize_names = TRUE,
  verbose = TRUE
)

Arguments

x

A datras_raw object.

vars_hh

Character vector of haul-level variables to keep from HH.

vars_hl

Character vector of variables from HL to spread wide across species. Defaults to "Count".

species_var

Character scalar giving the HL variable that defines the species-specific column names. Defaults to "Species". Can also be set to "Valid_Aphia" or another grouping variable present in HL.

id_var

Character scalar giving the haul identifier variable used to match HH and HL. Defaults to "haul.id".

sep

Character scalar used to separate the HL variable name from the species name in wide column names. Defaults to "__".

fill

Value used to replace missing values in the wide HL columns. Defaults to 0.

sanitize_names

Logical. If TRUE (default), species names used in wide column names are converted to lower case and non-alphanumeric characters are replaced with underscores.

verbose

Logical. If TRUE (default), warnings are issued for requested variables that are not found.

Value

A data frame in wide format with one row per haul.

Details

The output contains one row per haul. Requested haul-level variables from HH are kept as ordinary columns, while requested variables from HL are expanded into columns of the form "<variable><sep><species>".

The function starts from the HH table and adds requested variables from HL after aggregating them to unique haul.id x species_var combinations.

Numeric HL variables are summed within haul-species combinations before reshaping wide. Non-numeric variables are reduced by taking the first non-missing value.

Variables requested in vars_hh or vars_hl that are not found in the relevant table are omitted and reported with a warning.

See also

as_long_format()

Examples

if (FALSE) { # \dontrun{
## One row per haul, species-specific count columns
tab <- as_wide_format(x)

## Add more HL variables
tab <- as_wide_format(
  x,
  vars_hl = c("Count", "CatCatchWgt")
)

## Use Aphia IDs instead of species names in column names
tab <- as_wide_format(
  x,
  vars_hl = c("Count"),
  species_var = "Valid_Aphia"
)
} # }