which waveform viewer can be easily be integrated with c or c++?

Hello,

    I have a requirement, wherein my dsp datapath algorithms are
    in C or C++. The input and output data is stored in memory
    buffers. Now staying in C environment I want to see the plots
    of the waveforms. So which waveform viewer could easily be 
    integrated in C/c++ environment. 

    1. once the figure window pops up, I should be able to zoom
       into the plot.

    2. multiple figure windows should be supported.

    3. subplot should also be supported.

    I am looking for matlab plot kind of capabilities but more
    from free and open source softwares.

Regards
Bharat

On 08/10/2010 06:31 PM, bharat pathak wrote:
> Hello,
>
>      I have a requirement, wherein my dsp datapath algorithms are
>      in C or C++. The input and output data is stored in memory
>      buffers. Now staying in C environment I want to see the plots
>      of the waveforms. So which waveform viewer could easily be
>      integrated in C/c++ environment.
>
>      1. once the figure window pops up, I should be able to zoom
>         into the plot.
>
>      2. multiple figure windows should be supported.
>
>      3. subplot should also be supported.
>
>      I am looking for matlab plot kind of capabilities but more
>      from free and open source softwares.
>
> Regards
> Bharat

Check out Gnuplot.  I know almost nothing about it, other than it is 
used by Octave and a few other Gnu-ish scientific applications.

-- 

Tim Wescott
Wescott Design Services
http://www.wescottdesign.com

Do you need to implement control loops in software?
"Applied Control Theory for Embedded Systems" was written for you.
See details at http://www.wescottdesign.com/actfes/actfes.html

On Tue, 10 Aug 2010 20:31:28 -0500, "bharat pathak"
<bharat@n_o_s_p_a_m.arithos.com> wrote:

>Hello,
>
>    I have a requirement, wherein my dsp datapath algorithms are
>    in C or C++. The input and output data is stored in memory
>    buffers. Now staying in C environment I want to see the plots
>    of the waveforms. So which waveform viewer could easily be 
>    integrated in C/c++ environment. 
>
>    1. once the figure window pops up, I should be able to zoom
>       into the plot.
>
>    2. multiple figure windows should be supported.
>
>    3. subplot should also be supported.
>
>    I am looking for matlab plot kind of capabilities but more
>    from free and open source softwares.
>
>Regards
>Bharat

Gnuplot. Used by Octave to emulate matlab plotting functionality.
-- 
Muzaffer Kal

DSPIA INC.
ASIC/FPGA Design Services

http://www.dspia.com

>On Tue, 10 Aug 2010 20:31:28 -0500, "bharat pathak"
><bharat@n_o_s_p_a_m.arithos.com> wrote:
>
>>Hello,
>>
>>    I have a requirement, wherein my dsp datapath algorithms are
>>    in C or C++. The input and output data is stored in memory
>>    buffers. Now staying in C environment I want to see the plots
>>    of the waveforms. So which waveform viewer could easily be 
>>    integrated in C/c++ environment. 
>>
>>    1. once the figure window pops up, I should be able to zoom
>>       into the plot.
>>
>>    2. multiple figure windows should be supported.
>>
>>    3. subplot should also be supported.
>>
>>    I am looking for matlab plot kind of capabilities but more
>>    from free and open source softwares.
>>
>>Regards
>>Bharat
>
>Gnuplot. Used by Octave to emulate matlab plotting functionality.

Gnuplot is an excellent choice, but don't be put off by some of the
material you can find about it on the web. You could easily mistake it for
a Mickey Mouse function drawing program from a lot of the descriptions you
will find. Its actually a rather powerful plotting tool, with a lot of
flexibility for massaging the appearance of the images it produces.

Steve

steveu <steveu@n_o_s_p_a_m.coppice.org> wrote:

>Gnuplot is an excellent choice, but don't be put off by some of the
>material you can find about it on the web. You could easily mistake it for
>a Mickey Mouse function drawing program from a lot of the descriptions you
>will find. Its actually a rather powerful plotting tool, with a lot of
>flexibility for massaging the appearance of the images it produces.

I haven't used it, but I imagine it beats Excel, which I often
use for plotting.

(Excel has some advantages of its own, but I doubt they stack up.)

Tangentially, I reserve the term "waveform viewer" for tools that
can probe into an annotated simulation.  I do not think Gnuplot
can do that with C++, unless someone has gone to the effort of
merging it with a symbolic debugger.

Steve

On 11/08/10 05:27, Steve Pope wrote:
> steveu<steveu@n_o_s_p_a_m.coppice.org>  wrote:
>
>> Gnuplot is an excellent choice, but don't be put off by some of the
>> material you can find about it on the web. You could easily mistake it for
>> a Mickey Mouse function drawing program from a lot of the descriptions you
>> will find. Its actually a rather powerful plotting tool, with a lot of
>> flexibility for massaging the appearance of the images it produces.
>
> I haven't used it, but I imagine it beats Excel, which I often
> use for plotting.
>
> (Excel has some advantages of its own, but I doubt they stack up.)
>
> Tangentially, I reserve the term "waveform viewer" for tools that
> can probe into an annotated simulation.  I do not think Gnuplot
> can do that with C++, unless someone has gone to the effort of
> merging it with a symbolic debugger.
>
> Steve

Dislin is worth a look: : http://www.mps.mpg.de/dislin/

On 8/10/2010 8:31 PM, bharat pathak wrote:
>      I have a requirement, wherein my dsp datapath algorithms are
>      in C or C++. The input and output data is stored in memory
>      buffers. Now staying in C environment I want to see the plots
>      of the waveforms. So which waveform viewer could easily be
>      integrated in C/c++ environment.
>
>      1. once the figure window pops up, I should be able to zoom
>         into the plot.
>
>      2. multiple figure windows should be supported.
>
>      3. subplot should also be supported.
>
>      I am looking for matlab plot kind of capabilities but more
>      from free and open source softwares.

Hi Bharat,

FWIW, I've looked for this sort of thing over the years and never have 
found anything that's suitable for my products, so I ended up rolling my 
own.  It was a great deal of work, and evolved over a period of years. 
It's been beneficial, though, to get exactly what I want compared to 
making compromises with other solutions.

ScopeDSP began many years ago using a "VBX" plotting control provided by 
Visual Basic, which could be accessed by C++.  It was pretty good 
overall but it had an Achilles heal that it couldn't do bar plots with 
an arbitrary origin of the type typically used to display FFT results 
from an arbitrary base like -120 dB.  That's when I started to write my 
own.  That ended up being a good thing when Microsoft abandoned VBX 
controls a short time after hawking them as the next big thing.  (.net 
is next. ;-)

I've looked at gnuplot over the years, and it's very appealing in terms 
of being fairly easy to interface with via a pipe.  It also has a lot of 
features.  However, from a user's point of view, I always find its 
zooming to be a drag.  I always feel like I'm slumming when I have to 
use it from Octave.  (Not that Matlab's zooming is any better...)

Ideally, this sort of thing would be available from wxWidgets, QT, or 
other similar cross-platform GUI toolkits.  However, those don't seem to 
include a sophisticated plotting capability.  Maybe they think that's 
out of their scope, or more likely, nobody has ever contributed 
something like that.  So, I've thought of contributing my system, but 
since it's currently written in MFC, that would be a great deal of work. 
  So it hasn't happened.

Richard mentioned dislin; I think I've looked at that in the past but I 
can't remember why I didn't get too interested in it. I'll look at it 
again, though.

You might also look at Python, which has something called "matplotlib" 
as an add-on.  It's possible to mix C/C++ with Python.  Or, you could 
maybe transliterate matplotlib into C/C++.  That would probably be a lot 
of work, though.

Overall, I'd think that gnuplot would be the easiest thing, if you don't 
mind slumming a little. :-)

Grant

-- 
_____________________________________________________________________

Grant R. Griffin
Publisher of dspGuru                          http://www.dspguru.com
Iowegian International Corporation            http://www.iowegian.com
See http://www.iowegian.com/img/contact.gif for e-mail address

Thank you Grant and All,

Regards
Bharat

Grant Griffin <nobody@example.com> writes:
> On 8/10/2010 8:31 PM, bharat pathak wrote:
>> ...
>>      buffers. Now staying in C environment I want to see the plots
>>      of the waveforms. So which waveform viewer could easily be
>>      integrated in C/c++ environment.
>> ...
> ...
> FWIW, I've looked for this sort of thing over the years and never have
> found anything that's suitable for my products, so I ended up rolling

My experience too.

> ...
> Ideally, this sort of thing would be available from wxWidgets, QT, or
> other similar cross-platform GUI toolkits.  However, those don't seem
> to include a sophisticated plotting capability. ...

Did you look at GtkPlot from the GtkExtra widget set? 
(http://gtkextra.sourceforge.net/). IIRC there is also an OpenSource
app that uses these widgets for scientific plotting.

The GtkImageViewer widget (search engine) might also be of interest. 

>  Maybe they think
> that's out of their scope, or more likely, nobody has ever contributed
> something like that.  So, I've thought of contributing my system, but
> since it's currently written in MFC, that would be a great deal of
> work. So it hasn't happened.

Well, there is a rule in Open Source: release early and release
often. (Eric Raymond IIRC.) There might be people who are willing to
spend the time porting to, say, GTk+ and other open librarys.  Any
way, there are so powerfull tools for softwaredevelopment today, that
the biggest value of your app might be the ``design'': userinterface,
concepts, algorithms.  It might be a _big service_ to the community if
you wrote a paper where you expose this; describing what is intended
might be more interesting than what you implemented.  Kind of
blueprint.  Obviously that's the info people who are software
developers in the first place are looking for.

>
> Richard mentioned dislin; I think I've looked at that in the past but
> I can't remember why I didn't get too interested in it. I'll look at
> it again, though. ...

I looked at it last time about a year ago: written in FORTRAN --- at
least in the beginning --- and smells FORTRAN.  I wrote my first
programs in FORTRAN IV on punchcards, you know; I can't stand that
crude approach to programming any longer.

It's not only for the lack of data abstraction.  It's also the
monolithic approach.  E.g. I did not see that DISLIN makes use of
advanced rendering libraries such as Cairo or AGG as modern graphics
libs like Gtk+ and friends do or can.

-- 
hw

Heinrich Wolf wrote:
>Grant Griffin <nobody@example.com> writes:
>>  Maybe they think
>> that's out of their scope, or more likely, nobody has ever contributed
>> something like that.  So, I've thought of contributing my system, but
>> since it's currently written in MFC, that would be a great deal of
>> work. So it hasn't happened.
>
>Well, there is a rule in Open Source: release early and release
>often. (Eric Raymond IIRC.) There might be people who are willing to
>spend the time porting to, say, GTk+ and other open librarys.  Any
>way, there are so powerfull tools for softwaredevelopment today, that
>the biggest value of your app might be the ``design'': userinterface,
>concepts, algorithms.  It might be a _big service_ to the community if
>you wrote a paper where you expose this; describing what is intended
>might be more interesting than what you implemented.  Kind of
>blueprint.  Obviously that's the info people who are software
>developers in the first place are looking for.


And there ARE people who still use MFC.  Even if you don't have time, some
would probably still like to see it as-is. :)

Michael