[detail]
I have data with 4 colums: x, y, z, label.
'label' is integer and increases one by one.
I would like to shuffle 'label' when it is plotted:
sp 'file' u 1:2:3:4 lc palette
Thanks,
masa
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
What I want to do is following (let me clarify my question by 2D plot):
prepare 'data.txt':
1.0 2.0 1
2.0 3.0 1
3.0 4.0 1
1.0 2.5 2
2.0 3.5 2
3.0 4.5 2
1.0 4.5 3
2.0 3.5 3
3.0 2.5 3
The following command plot as it is:
pl 'data.txt' u 1:2:3 w lp lc palette
I want to do something shuffling (switching) the 3rd column randomly,
so I plot the data as
1.0 2.0 1*rand() (=3?)
2.0 3.0 1*rand() (=3?)
3.0 4.0 1*rand() (=3?)
Sorry, I do not understand.
Coloring by palette implies a continuous variable between 0.0 and 1.0
Coloring by discrete integers is something else entirely.
If you want a random palette color for the data points given above, then use
plot 'data.txt' using 1:2:(rand(0)) with points pointsize 5 lc palette
Using randomly colored lines is usually confusing, but points work OK.
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
What I attached is a figure produced by
splot 'data.txt' using 1:2:3:4 w l lc palette
I would like to change the color of rings randomly.
Is there a way to do that in inside the gnuplot?
Or do I need to shuffle the label when I produce data file?
p.s.1
I understand palette is for continuous color changes.
But in my case, the number of rings are huge and the labeling is almost continuous in other data.
p.s.2
plot 'data.txt' using 1:2:(rand(0)) with points pointsize 5
gives me the same color plot. Is it correct?
I'm using newest version of gnuplot.
Oops. I clicked the "delete" button rather than the "download" button, so I lost your pdf plot.
Still not sure I understand, maybe this will help...
The "index" keyword selects one block of data from your file. The index number is (I think) the same thing you have in column 4 of the data. So this command should give you a distinct line color for each indexed block:
splot for [i=1:*] 'data2.txt' index i using 1:2:3 lc i
This is not palette coloring. It uses the normal sequence of colors by linetype. The iteration using a star symbol to mean "until end of file" is only supported in the CVS version of gnuplot, so I hope that is what you mean by "newest". Otherwise you could say for [i=1:999] but you may get an error message for every iteration with i beyond the end of the file.
To add randomness you could replace "lc i" with some other constant color. Do not use the 'palette' keyword because that is not a constant. Maybe something like this to generate RGB colors with random components R G B.
RGB(r,g,b) = sprintf("#%06x",256.*256.*255.*r+256.*255.*g+255.*b)
splot for [i=1:*] 'data2.txt' index i using 1:2:3 lc rgb RGB(rand(0), rand(0), rand(0))
If you would like to refer to this comment somewhere else in this project, copy and paste the following link:
Hi,
I want to randomly change color of the plot.
In MATLAB I found using label2rgb with option 'shuffle'
http://blogs.mathworks.com/steve/2012/05/15/shuffling-label-colors/
Is there a way to do it in gnuplot?
[detail]
I have data with 4 colums: x, y, z, label.
'label' is integer and increases one by one.
I would like to shuffle 'label' when it is plotted:
sp 'file' u 1:2:3:4 lc palette
Thanks,
masa
I don't understand what you mean by "shuffle", and I don't know what the meaning of "label" is. But...
If you want to use a random palette color for each point you can do:
Thanks for your reply.
What I want to do is following (let me clarify my question by 2D plot):
prepare 'data.txt':
1.0 2.0 1
2.0 3.0 1
3.0 4.0 1
1.0 2.5 2
2.0 3.5 2
3.0 4.5 2
1.0 4.5 3
2.0 3.5 3
3.0 2.5 3
The following command plot as it is:
pl 'data.txt' u 1:2:3 w lp lc palette
I want to do something shuffling (switching) the 3rd column randomly,
so I plot the data as
1.0 2.0 1*rand() (=3?)
2.0 3.0 1*rand() (=3?)
3.0 4.0 1*rand() (=3?)
1.0 2.5 2*rand() (=1?)
2.0 3.5 2*rand() (=1?)
3.0 4.5 2*rand() (=1?)
1.0 4.5 3*rand() (=2?)
2.0 3.5 3*rand() (=2?)
3.0 2.5 3*rand() (=2?)
I want to do something like the 3rd and 4th plot of following web site:
http://blogs.mathworks.com/steve/2012/05/15/shuffling-label-colors/
Is it possible?
Last edit: Masahiro Takahashi 2016-02-03
Sorry, I do not understand.
Coloring by palette implies a continuous variable between 0.0 and 1.0
Coloring by discrete integers is something else entirely.
If you want a random palette color for the data points given above, then use
Using randomly colored lines is usually confusing, but points work OK.
Ummm, sorry for my bad explanation.
What I attached is a figure produced by
splot 'data.txt' using 1:2:3:4 w l lc palette
I would like to change the color of rings randomly.
Is there a way to do that in inside the gnuplot?
Or do I need to shuffle the label when I produce data file?
p.s.1
I understand palette is for continuous color changes.
But in my case, the number of rings are huge and the labeling is almost continuous in other data.
p.s.2
plot 'data.txt' using 1:2:(rand(0)) with points pointsize 5
gives me the same color plot. Is it correct?
I'm using newest version of gnuplot.
Last edit: Masahiro Takahashi 2016-02-03
Oops. I clicked the "delete" button rather than the "download" button, so I lost your pdf plot.
Still not sure I understand, maybe this will help...
The "index" keyword selects one block of data from your file. The index number is (I think) the same thing you have in column 4 of the data. So this command should give you a distinct line color for each indexed block:
This is not palette coloring. It uses the normal sequence of colors by linetype. The iteration using a star symbol to mean "until end of file" is only supported in the CVS version of gnuplot, so I hope that is what you mean by "newest". Otherwise you could say for [i=1:999] but you may get an error message for every iteration with i beyond the end of the file.
To add randomness you could replace "lc i" with some other constant color. Do not use the 'palette' keyword because that is not a constant. Maybe something like this to generate RGB colors with random components R G B.
RGB(r,g,b) = sprintf("#%06x",256.*256.*255.*r+256.*255.*g+255.*b) splot for [i=1:*] 'data2.txt' index i using 1:2:3 lc rgb RGB(rand(0), rand(0), rand(0))I thought of a way to use random colors drawn specifically from the palette
This works because the color definition is executed only once per iteration, and the resulting value is then treated as a constant palette color
Sorry for my replay.
Thanks for your answer.
This is what I wanted to do!
Thank you!
Ummm, I don't understand either
Sorry my bad explanation...