Re: [Gnuplot-info] Multiple sources on a common X axis

classic Classic list List threaded Threaded
3 messages Options
Reply | Threaded
Open this post in threaded view
|

Re: [Gnuplot-info] Multiple sources on a common X axis

Hans-Bernhard Bröker-2
Am 29.07.2016 um 09:12 schrieb Robert von Knobloch:

> I _think_ that I _do_ want multiplot.

Then gnuplot cannot figure out the shared axis range all by itself, like
you said you wanted.  It's a case of either having the cake, or eating it.

The individual plots in a multiplot are done independently, so there's
just no way for the x data range of later plots to have any influence on
the x axis range of earlier ones, because by the time the second plot is
made, the first one is completely done.  No going back to make its x
axis range match the latter ones.

> I want the temperature plot to be framed in its own axes, just on the
> same page (screen), with the same x-axis (date/time) as the other data.
> My screen has then 3 (or more with each having different Y units) plots
> with their own y axes, but with the same x scale and labels, one above
> the other, so that it is easy to see correlations.

In that case you'll have to decide on an x axis range beforehand, tell
gnuplot about your choice, fix the left and right margins, and then you
can use
       
        set multiplot layout 3, 1

But for plots with at most two incommensurable types of y data, a
single, multi-dataset plot with a secondary y axis will make better use
of the available plotting surface, and allow for a much clearer view of
correlations, and it'll be a whole lot easier to create.

------------------------------------------------------------------------------
_______________________________________________
gnuplot-beta mailing list
[hidden email]
Membership management via: https://lists.sourceforge.net/lists/listinfo/gnuplot-beta
Reply | Threaded
Open this post in threaded view
|

Re: [Gnuplot-info] Multiple sources on a common X axis

Robert von Knobloch
On 29/07/16 21:23, Hans-Bernhard Bröker wrote:

> In that case you'll have to decide on an x axis range beforehand, tell
> gnuplot about your choice, fix the left and right margins, and then you
> can use
>
> set multiplot layout 3, 1
>
> But for plots with at most two incommensurable types of y data, a
> single, multi-dataset plot with a secondary y axis will make better use
> of the available plotting surface, and allow for a much clearer view of
> correlations, and it'll be a whole lot easier to create.

This is my problem.
I cannot fix the x-axis (it grows).
I have 5 or 6 different Y-data so a plot on one set of axes looks a real
mess.
Bob

------------------------------------------------------------------------------
_______________________________________________
gnuplot-beta mailing list
[hidden email]
Membership management via: https://lists.sourceforge.net/lists/listinfo/gnuplot-beta
Reply | Threaded
Open this post in threaded view
|

Re: [Gnuplot-info] Multiple sources on a common X axis

Hans-Bernhard Bröker-2
Am 01.08.2016 um 10:12 schrieb Robert von Knobloch:
> On 29/07/16 21:23, Hans-Bernhard Bröker wrote:
>> In that case you'll have to decide on an x axis range beforehand, tell
>> gnuplot about your choice,

> I cannot fix the x-axis (it grows).

I didn's say you have to fix it.  But _someone_ has to decide what the
common axis range for all those data sets is going to be.  Since you
defined the boundary conditions of the job such there's no way left for
gnuplot to do it, that someone has to be you, or some program/script you
write for the job.

gnuplot can only do it across one plot's collection of data sets --- but
you're dead set on having a multi-plot, so that's out.

And there's no way gnuplot can guess what the data in the 5th plot in a
mulit-plot will look like, to scale the first plot accordingly, so
gnuplot doing it automatically for the multiplot is clearly out, too.

So _you_ have to make the decision, before you issue the first plot
command.  How you arrive at the range to use is for you to figure out.
The 'stat' command might help, if the data sets remain stable for at
least as long as the entire command multiplot sequence takes.  Or maybe
the programs generating those files could put that knowledge somewhere
for gnuplot to get at.  You may have to make snapshot copies of the data
files for the duration.



------------------------------------------------------------------------------
_______________________________________________
gnuplot-beta mailing list
[hidden email]
Membership management via: https://lists.sourceforge.net/lists/listinfo/gnuplot-beta