How to plot a multicolum file in gnuplot - colors

Question

I am having a multicoulum file and after certain rows it has a row break.
It looks like
# SET: 1
0.00000 -62.49368 0.07000
0.00639 -62.49367 0.07000
0.01276 -62.49367 0.07000
0.01914 -62.49366 0.07000
0.02553 -62.49365 0.07000
0.03190 -62.49364 0.07000
0.03829 -62.49362 0.07000
0.04467 -62.49361 0.07000
0.05106 -62.49359 0.07000
0.05743 -62.49356 0.07000
# SET: 2
0.00000 -62.49342 0.07000
0.00639 -62.49342 0.07000
0.01276 -62.49341 0.07000
0.01914 -62.49340 0.07000
0.02553 -62.49339 0.07000
0.03190 -62.49338 0.07000
0.03829 -62.49337 0.07000
0.04467 -62.49335 0.07000
# SET: 3
0.00000 -62.47334 0.07000
0.00639 -62.47225 0.07000
0.01276 -62.47228 0.07000
0.01914 -62.47231 0.07000
0.02553 -62.47236 0.07000
0.03190 -62.47242 0.07000
0.03829 -62.47248 0.07000
0.04467 -62.47256 0.07000
0.05106 -62.47264 0.07000
0.05743 -62.47273 0.07000
0.06381 -62.47283 0.07000
0.07020 -62.47296 0.07000
0.07657 -62.47296 0.07000
and so on.
My plot is produced using
plot 'data' u 1:2 w l lc 2
The y-axis data are around zero from some negative to positive scale alining along x-axis.
I want to make top line just below the zero and just above the zero (horizontal line at Y=0 at) with different colour.
The data set which is to be plotted near to Y=0 (along x-axis) from data file looks like ($2 will flip the sign from negative to positive)
# SET: 38
0.00000 -0.88752 0.07000
0.00639 -0.88731 0.07000
0.01276 -0.88751 0.07000
0.01914 -0.88783 0.07000
0.02553 -0.88827 0.07000
0.03190 -0.88884 0.07000
0.03829 -0.88954 0.07000
0.04467 -0.89036 0.07000
0.05106 -0.89132 0.07000
0.05743 -0.89240 0.07000
# SET: 39
0.00000 2.02394 0.07000
0.00639 2.02456 0.07000
0.01276 2.02642 0.07000
0.01914 2.02950 0.07000
0.02553 2.03379 0.07000
0.03190 2.03927 0.07000
0.03829 2.04590 0.07000
0.04467 2.05364 0.07000
0.05106 2.06264 0.07000
0.05743 2.07248 0.07000
0.06381 2.08330 0.07000
0.07020 2.09494 0.07000
0.07657 2.10755 0.07000
So basically I want to make the both the lines colorful which changes the sign at colum 2 from negative to positive. The line can be grepped using awk and grep commands and can be seen that after which SET $2 changes from negative to positive and then that SET index can be used in gnuplot to make the line colorful.
Unfortunately I do not know how to do it.
Hope I have explained it nicely. Please let me know if I need to clarify it more.

Answer 1

If I understand you correctly you want to have the line in one color (e.g. red 0xff0000) when the value in column 2 is smaller zero and and another color (e.g. green 0x00ff00) if it is larger zero. Simply define a function which returns a color code in the form of 0xRRGGBB.
Maybe something like this?
(If you are plotting from file skip the part $Data <<EOD ... EOD and in the plot command replace $Data with your filename, e.g. 'MyFile.dat'.
Code:
### color dependent on column value
reset session
$Data <<EOD
# SET: 38
0.00000 -0.88752 0.07000
0.00639 -0.88731 0.07000
0.01276 -0.88751 0.07000
0.01914 -0.88783 0.07000
0.02553 -0.88827 0.07000
0.03190 -0.88884 0.07000
0.03829 -0.88954 0.07000
0.04467 -0.89036 0.07000
0.05106 -0.89132 0.07000
0.05743 -0.89240 0.07000
# SET: 39
0.00000 2.02394 0.07000
0.00639 2.02456 0.07000
0.01276 2.02642 0.07000
0.01914 2.02950 0.07000
0.02553 2.03379 0.07000
0.03190 2.03927 0.07000
0.03829 2.04590 0.07000
0.04467 2.05364 0.07000
0.05106 2.06264 0.07000
0.05743 2.07248 0.07000
0.06381 2.08330 0.07000
0.07020 2.09494 0.07000
0.07657 2.10755 0.07000
EOD
myColor(n) = n<0 ? 0xff0000 : 0x00ff00
plot $Data u 1:2:(myColor($2)) w lp pt 7 lc rgb var notitle
### end of code
Result:
Addition:
After clarification in the comments the following might be a possible solution.
You plot your data into a dummy table and check which is the last set which has column2-values smaller than zero. column(-1) contains the number of the set (counting starts from zero). Check also help pseudocolumns.
You modify your color function a bit and take the colors from an array.
Tested with gnuplot 5.2.6.
Code:
### color dependent on column value
reset session
# create some random test data
set print $Data
Sets = 30
GapPos = int(rand(0)*Sets/3)+Sets/3
GapSize = 3
f(x) = (x-GapPos)+rand(0)*0.5 + (x>=GapPos ? GapSize : 0)
do for [i=1:Sets] {
print "# SET: ".i
do for [j=1:10] {
print sprintf("%.4f %.4f", (j+rand(0))/10., f(i))
}
print "\n"
}
set print
# find the SET-no. where column 2 values change from negative to positive
set table $Dummy
plot $Data u (SetNo = $2<0 ? column(-1)+1 : SetNo) with table
unset table
# set the colors
array ArrColors[6] = [0xff0000, 0x00ff00, 0x0000ff, 0xff00ff, 0xffff00, 0x00ffff]
myColor(n) = n<SetNo-3 || n>=SetNo+3? 0xcccccc : ArrColors[n-SetNo+4]
plot $Data u 1:2:(myColor(column(-1))) w lp pt 7 lc rgb var notitle
### end of code
Result: