I found many solutions to this question for GNUPlot 4.2+ using title column, columheader etc. However, I am stuck with GNUPlot 4.0 and need to find a way to automatically set the titles using the top row of my data set.
I'm using a bash script to generate a dynamic (-> changes daily) file with data and I have a static .gnu file that contains the gnuplot settings.
Do you have any suggestions?
(Perhaps a way to use a variable/array inside the .gnu file?)
The trick is using the Command Substitution
# content of file
$ cat data.txt
header
1
2
3
4
5
# generate gnuplot script
$ echo "set term dumb; plot '<(tail -n +2 data.txt)' title '$(head -1 data.txt)'"
set term dumb; plot '<(tail -n +2 data.txt)' title 'header'
# pipe to gnuplot
$ echo "set term dumb; plot '<(tail -n +2 data.txt)' title '$(head -1 data.txt)'" | gnuplot
# or you can use `here-doc` (note: `>` is bash prompt)
$ gnuplot <<_EOF_
> set term dumb
> plot '<(tail -n +2 data.txt)' title '$(head -1 data.txt)'
>_EOF_
5 ++-------+--------+-------+--------+--------+--------+-------+-------+A
+ + + + + + + header A +
4.5 ++ ++
| |
| |
4 ++ A ++
| |
3.5 ++ ++
| |
| |
3 ++ A ++
| |
2.5 ++ ++
| |
| |
2 ++ A ++
| |
1.5 ++ ++
| |
+ + + + + + + + +
1 A+-------+--------+-------+--------+--------+--------+-------+-------++
0 0.5 1 1.5 2 2.5 3 3.5 4
Note: $(head -1 data.txt) expand to header before passing to Gnuplot.