|
# This example shows how a nonlinear axis definition can be used to
# set up "probability axis" scaling on both the x and y axes.
# This is also known as "probit" scaling.
# We compare 3 treatments of an image file (Tux).
# They differ in the sigma value defining the normal distribution curve
# used for scaling.
# The 4th plot shows the original, unscaled image using linear axes.
#
save_encoding_3 = GPVAL_ENCODING
set encoding utf8
unset border
set title offset 0,-1
set multiplot layout 2,2 title "Probability axes: Scale image pixels by distance from center treated as a Z-score"
set title "Sigma = 22 pixels"
sigma = 22.
ticinterval = 2
load 'probably_tux.dem'
set title "Sigma = 30 pixels"
sigma = 30.
ticinterval = 1
load 'probably_tux.dem'
set title "Sigma = 60 pixels"
sigma = 60.
load 'probably_tux.dem'
set title "Linear Scale"
set label 1 "Tux"
set xlabel " "
set ylabel " "
unset nonlinear x
unset nonlinear y
unset xtics
unset ytics
replot
unset multiplot
Click here for minimal script to generate this plot
|