Skip to contents

Collapse a spectral cube into a white-light image (robust, weighted)

Source: R/collapse.R
collapse_white_light.Rd

Steps:

  1. subtract spatial median per-wavelength,

  2. clip very negative outliers at -k * MAD,

  3. (optional) inverse-variance weights 1/MAD^2,

  4. weighted sum and reshape to a 2-D image,

  5. force non-negative for visualization safety.

Usage

collapse_white_light(cube, kclip = 2, use_weights = TRUE)

Arguments

cube

3-D numeric array with dimensions [nx, ny, nlambda].

kclip

numeric; negative clip in units of MAD (default 2)

use_weights

logical; if TRUE, use 1/MAD^2 weights (default TRUE)

Value

2-D matrix with dimensions [nx, ny].