Skip to contents

Performs data transformation on a variable based on the specified method.

Source: R/adm_trasform.R
adm_transform.Rd

This function transforms the data in a tibble or SpatRaster object on the method specified The available methods are "01", "zscore", "log", and "round".

Usage

adm_transform(data, variable, method, inverse = FALSE, t_terms = NULL)

Arguments

data

A data.frame, tibble, or SpatRaster containing the data.

variable

A character string specifying the variable (column) to be transformed.

method

A character string specifying the method to be used for transformation. Available methods are "01", "zscore", "log", and "round."

  • For "01", it scales the variable between 0 and 1 using the formula (x - min(x)) (max(x) - min(x)).

  • For "zscore", it standardizes the variable by subtracting the mean and dividing by the standard deviation.

  • For "log", it applies natural logarithm transformation to the variable.

  • For "log1", it sums 1 and then applies natural logarithm transformation to the variable.

  • For "round", it rounds the variable's values to the nearest whole numbers.

inverse

logical. Invert the transformation?

t_terms

vector. c(a,b):

  • For "01", a = min(x), b = max(x).

  • For "zscore", a = mean(x), b = sd(x).

  • For "log" and "log1, not needed.

  • Can't invert "round" transformations.

Value

A data.frame or tibble with the transformed variable added as a new column. The new column's name is the original variable name followed by an underscore and method name.

Examples

if (FALSE) {
require(dplyr)
require(terra)

# Select data for a single species
data("sppabund")
some_sp <- sppabund %>%
  dplyr::filter(species == "Species one") %>%
  dplyr::select(species, ind_ha, x, y)

envar <- system.file("external/envar.tif", package = "adm")
envar <- terra::rast(envar)[["bio12"]]

# Transform tabular data
## Transform abundance data to 0-1
some_sp_2 <- adm_transform(
  data = some_sp,
  variable = "ind_ha",
  method = "01"
)
some_sp_2

## Transform abundance data z-score
some_sp_2 <- adm_transform(
  data = some_sp,
  variable = "ind_ha",
  method = "zscore"
)
some_sp_2

## Transform abundance data log
some_sp_2 <- adm_transform(
  data = some_sp,
  variable = "ind_ha",
  method = "log"
)
some_sp_2

## Round abundance data
some_sp_2 <- adm_transform(
  data = some_sp,
  variable = "ind_ha",
  method = "round"
)
some_sp_2

# Tranform raster data
## Transform to 0-1
envar_2 <- adm_transform(
  data = envar,
  variable = "bio12",
  method = "01"
)
envar_2

## Transform z-score
envar_2 <- adm_transform(
  data = envar,
  variable = "bio12",
  method = "zscore"
)
envar_2

## Transform log
envar_2 <- adm_transform(
  data = envar,
  variable = "bio12",
  method = "log"
)
envar_2

## Round
envar_2 <- adm_transform(
  data = envar,
  variable = "bio12",
  method = "round"
)
envar_2

# Invert transformation
## Invert 01 tranformation
some_sp_transformed <- adm_transform(
  data = some_sp,
  variable = "ind_ha",
  method = "01"
)
some_sp_transformed

some_sp_inverted <- adm_transform(
  data = some_sp_transformed,
  variable = "ind_ha_01",
  method = "01",
  inverse = TRUE,
  t_terms = c(
    a = min(some_sp[["ind_ha"]]),
    b = max(some_sp[["ind_ha"]])
  )
)
some_sp_inverted

## Invert z-score tranformation
some_sp_transformed <- adm_transform(
  data = some_sp,
  variable = "ind_ha",
  method = "zscore"
)
some_sp_transformed

some_sp_inverted <- adm_transform(
  data = some_sp_transformed,
  variable = "ind_ha_zscore",
  method = "zscore",
  inverse = TRUE,
  t_terms = c(
    a = mean(some_sp[["ind_ha"]]),
    b = sd(some_sp[["ind_ha"]])
  )
)
some_sp_inverted

## Invert log and log1
some_sp_transformed <- adm_transform(
  data = some_sp,
  variable = "ind_ha",
  method = "log"
)
some_sp_transformed

some_sp_inverted <- adm_transform(
  data = some_sp_transformed,
  variable = "ind_ha_log",
  method = "log",
  inverse = TRUE
)
some_sp_inverted

some_sp_transformed <- adm_transform(
  data = some_sp,
  variable = "ind_ha",
  method = "log1"
)
some_sp_transformed

some_sp_inverted <- adm_transform(
  data = some_sp_transformed,
  variable = "ind_ha_log1",
  method = "log1",
  inverse = TRUE
)
some_sp_inverted

## To invert raster, is the same process
## Example:
envar_transformed <- adm_transform(
  data = envar,
  variable = "bio12",
  method = "01"
)
envar_transformed

envar_inverted <- adm_transform(
  data = envar_transformed,
  variable = "bio12_01",
  method = "01",
  inverse = TRUE,
  t_terms = c(
    a = terra::global(envar, min, na.rm = T),
    b = terra::global(envar, max, na.rm = T)
  )
)
envar_inverted
}