Assess the appropriate raster layer data type

When writing raster-type objects to disk, a datatype can be specified. These functions help identify what smallest datatype can be used.

assessDataType(ras, type = "writeRaster")

# Default S3 method
assessDataType(ras, type = "writeRaster")

Arguments

ras

The RasterLayer or RasterStack for which data type will be assessed.

type

Character. "writeRaster" (default) or "GDAL" (defunct) to return the recommended data type for writing from the raster packages, respectively, or "projectRaster" to return recommended resampling type.

Value

A character string indicating the data type of the spatial layer (e.g., "INT2U"). See terra::datatype()

Examples

if (requireNamespace("terra", quietly = TRUE)) {
  ## LOG1S
  rasOrig <- terra::rast(ncols = 10, nrows = 10)
  ras <- rasOrig
  ras[] <- rep(c(0,1),50)
  assessDataType(ras)

  ras <- rasOrig
  ras[] <- rep(c(0,1),50)
  assessDataType(ras)

  ras[] <- rep(c(TRUE,FALSE),50)
  assessDataType(ras)

  ras[] <- c(NA, NA, rep(c(0,1),49))
  assessDataType(ras)

  ras <- rasOrig
  ras[] <- c(0, NaN, rep(c(0,1),49))
  assessDataType(ras)


  ## INT1S
  ras[] <- -1:98
  assessDataType(ras)

  ras[] <- c(NA, -1:97)
  assessDataType(ras)

  ## INT1U
  ras <- rasOrig
  ras[] <- 1:100
  assessDataType(ras)

  ras[] <- c(NA, 2:100)
  assessDataType(ras)

  ## INT2U
  ras <- rasOrig
  ras[] <- round(runif(100, min = 64000, max = 65000))
  assessDataType(ras)

  ## INT2S
  ras <- rasOrig
  ras[] <- round(runif(100, min = -32767, max = 32767))
  assessDataType(ras)

  ras[54] <- NA
  assessDataType(ras)

  ## INT4U
  ras <- rasOrig
  ras[] <- round(runif(100, min = 0, max = 500000000))
  assessDataType(ras)

  ras[14] <- NA
  assessDataType(ras)

  ## INT4S
  ras <- rasOrig
  ras[] <- round(runif(100, min = -200000000, max = 200000000))
  assessDataType(ras)

  ras[14] <- NA
  assessDataType(ras)

  ## FLT4S
  ras <- rasOrig
  ras[] <- runif(100, min = -10, max = 87)
  assessDataType(ras)

  ras <- rasOrig
  ras[] <- round(runif(100, min = -3.4e+26, max = 3.4e+28))
  assessDataType(ras)

  ras <- rasOrig
  ras[] <- round(runif(100, min = 3.4e+26, max = 3.4e+28))
  assessDataType(ras)

  ras <- rasOrig
  ras[] <- round(runif(100, min = -3.4e+26, max = -1))
  assessDataType(ras)

  ## FLT8S
  ras <- rasOrig
  ras[] <- c(-Inf, 1, rep(c(0,1),49))
  assessDataType(ras)

  ras <- rasOrig
  ras[] <- c(Inf, 1, rep(c(0,1),49))
  assessDataType(ras)

  ras <- rasOrig
  ras[] <- round(runif(100, min = -1.7e+30, max = 1.7e+308))
  assessDataType(ras)

  ras <- rasOrig
  ras[] <- round(runif(100, min = 1.7e+30, max = 1.7e+308))
  assessDataType(ras)

  ras <- rasOrig
  ras[] <- round(runif(100, min = -1.7e+308, max = -1))
  assessDataType(ras)

  # 2 layer with different types LOG1S and FLT8S
  ras <- rasOrig
  ras[] <- rep(c(0,1),50)
  ras1 <- rasOrig
  ras1[] <- round(runif(100, min = -1.7e+308, max = -1))
  sta <- c(ras, ras1)
  assessDataType(sta)

}
#> [1] "LOG1S" "FLT8S"