I want to plot a 18x18 matrix with gnuplot. Here is my codes:
set size ratio 1
set palette gray negative
set xrange[-0.5:17.5]
set yrange[-0.5:17.5]
set cbrange[-0.2:0.8]
set xtics 0,1,17
set ytics 0,1,17
set xtics offset -0.5,0
set title "Resolusition Matrix for E"
plot "Mat" matrix w image noti
Then I got a fig like this:
Now I would like to add borders to each cell, which will look like this:
Thank you.
For your case you can set one minor tic, which then lies on the border between two pixels, and draw a grid on them:
set size ratio 1
set palette gray negative
set autoscale xfix
set autoscale yfix
set xtics 1
set ytics 1
set title "Resolution Matrix for E"
set tics scale 0,0.001
set mxtics 2
set mytics 2
set grid front mxtics mytics lw 1.5 lt -1 lc rgb 'white'
plot "Mat" matrix w image noti
Note, that set grid front
also brings the tics to the front. To avoid that you can scale the tics to 0
. For the minor tics you must use a very small number, 0
would omit the grid lines on the minor tics.
The result with 4.6.3 is:
EDIT: In order to control the grid lines and tic labels independently, you can use the unused x2
and y2
to draw the grid (inspired by an answer to How do I draw a vertical line in gnuplot?):
set size ratio 1
set palette gray negative
# grid lines
set x2tics 1 format '' scale 0,0.001
set y2tics 1 format '' scale 0,0.001
set mx2tics 2
set my2tics 2
# labeling
set xtics 5 out nomirror
set ytics 5 out nomirror
set grid front mx2tics my2tics lw 1.5 lt -1 lc rgb 'white'
set xrange[-0.5:39.5]
set yrange[-0.5:39.5]
set x2range[-0.5:39.5]
set y2range[-0.5:39.5]
plot "Mat" matrix w image notitle
With gnuplot
version 4.6, this requires setting explicit ranges, so that the x
and x2
(unused!) are equal. The information might be extracted with stats
from the data file.
Using version 5 allows you to use set link
. Instead of all the set *range
stuff. You could use:
set autoscale fix
set link x
set link y
Result:
If you love us? You can donate to us via Paypal or buy me a coffee so we can maintain and grow! Thank you!
Donate Us With