interpolate.classes {EMMAgeo}

Interpolate data between different classes.

Description

This function interpolates grain-size data for different classes, either to higher or to lower resolution.

Usage

interpolate.classes(
  X,
  boundaries.in,
  boundaries.out,
  method = "natural",
  fixed.start = TRUE
)

Arguments

X

Numeric matrix, input data set with m samples (rows) and n variables (columns).

boundaries.in

Numeric vector, class boundaries of the input data.

boundaries.out

Numeric vector, class boundaries of the output data.

method

Logical scalar, interpolation method, one out of "linear" (linear interpolation), "fmm" (cubic spline), "natural" (natural spline), "periodic" (periodic spline). Default is "natural".

fixed.start

Logocal scalar, specifying if the outer boundaries should be set to the same values as in the original matrix, default is TRUE. This may become necessary to avoid interpolation errors, see example.

Value

Numeric matrix, interpolated class values.

Author(s)

Michael Dietze, Elisabeth Dietze

See Also

EMMA, approx, spline

Examples

## load example data
data(example_X)
classes.in <- seq(from = 1, to = 10, length.out = ncol(X))
  
## Example 1 - decrease the class numbers
## define number of output classes
classes.out <- seq(1, 10, length.out = 20)

## interpolate the data set
Y <- interpolate.classes(X = X, 
                         boundaries.in = classes.in, 
                         boundaries.out = classes.out,
                         method = "linear")

## show original vs. interpolation for first 10 samples
plot(NA, xlim = c(1, 10), ylim = c(0, 40))
for(i in 1:10) {
  lines(classes.in, X[i,] * 20 + i)
  lines(classes.out, Y[i,] * 20 + i, col = 2)
}

## Example 2 - increase the class numbers
## define number of output classes
classes.out <- seq(1, 10, length.out = 200)

## interpolate the data set
Y <- interpolate.classes(X = X, 
                         boundaries.in = classes.in, 
                         boundaries.out = classes.out)

## show original vs. interpolation for first 10 samples
plot(NA, xlim = c(1, 10), ylim = c(0, 40))
for(i in 1:10) {
  lines(classes.in, X[i,] * 20 + i)
  lines(classes.out, Y[i,] * 20 + i, col = 2)
}