Request for running compare_plot_demos

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

Request for running compare_plot_demos

Rik-4
Bug #48519 which prevented running all of the plots with the qt toolkit has
been fixed.

Could someone run compare_plot_demos with qt, fltk, and Matlab and then
report where to find the results?

--Rik

Reply | Threaded
Open this post in threaded view
|

Re: Request for running compare_plot_demos

siko1056
On Sat, Jan 12, 2019 at 3:01 AM Rik <[hidden email]> wrote:
Bug #48519 which prevented running all of the plots with the qt toolkit has
been fixed.

Could someone run compare_plot_demos with qt, fltk, and Matlab and then
report where to find the results?

--Rik


Thank you for taking so much time and effort on these bugs!  Here some small contribution to this.  It really took a while (about 2 Hours) but here we go (openSUSE 15.0; Octave 5.0.1 [hg-id_6657213145e5] with gnuplot, fltk, qt, and Matlab R2018b):


The Matlab script "dump_plot_demos" generated by "compare_plot_demos" is 17k lines long and Matlab refuses to run it:

   Error: File: dump_plot_demos.m Line: 1 Column: 1
   The input was too complicated or too big for MATLAB to parse.

Thus I needed to comment the halve, ran it, commented the other halve, ran it again, and finally merged the diary-outputs.
Additionally, four times the double parenthesis indexing ()() was used, which is not supported by Matlab. I split those occurrences in two statements.

Anyways, the results look very convincing!  Can we put this on the website after the release maybe? ;-)

Best,
Kai
Reply | Threaded
Open this post in threaded view
|

Re: Request for running compare_plot_demos

Rik-4
On 01/15/2019 04:10 AM, Kai Torben Ohlhus wrote:

Thank you for taking so much time and effort on these bugs!  Here some small contribution to this.  It really took a while (about 2 Hours) but here we go (openSUSE 15.0; Octave 5.0.1 [hg-id_6657213145e5] with gnuplot, fltk, qt, and Matlab R2018b):


The Matlab script "dump_plot_demos" generated by "compare_plot_demos" is 17k lines long and Matlab refuses to run it:

   Error: File: dump_plot_demos.m Line: 1 Column: 1
   The input was too complicated or too big for MATLAB to parse.

I love it.  Octave has bragging rights that it can run a script that Matlab can't even parse.


Thus I needed to comment the halve, ran it, commented the other halve, ran it again, and finally merged the diary-outputs.
Additionally, four times the double parenthesis indexing ()() was used, which is not supported by Matlab. I split those occurrences in two statements.

Can you report those four occurrences as a bug?  There's no reason we can't right the %!demo code to run under both Octave and Matlab which would make this easier in the future.



Anyways, the results look very convincing!  Can we put this on the website after the release maybe? ;-)


You're correct that we rarely do enough marketing on behalf of Octave.  Maybe we can do more to promote this release which looks to be a significant one.

--Rik


Reply | Threaded
Open this post in threaded view
|

Re: Request for running compare_plot_demos

mmuetzel
In reply to this post by siko1056
Am Dienstag, 15. Januar 2019 um 13:10 Uhr "Kai Torben Ohlhus" wrote:
> Anyways, the results look very convincing!  Can we put this on the website after the release maybe? ;-)
 
Thanks for running these tests.
You're right: I also think that these are the best results we had so far.

Just a few observations without looking at any of the demo code:
- Something seems to be off with the camzoom demo and the qt toolkit.
- The syntax for our copyobj doesn't seem to be Matlab compatible.
- Either ours or Matlab's plotboxaspectratio seems to be off. E.g. for demo ellipsoid.
- We might be doing something wrong with the z axis scale for zoom...
- And finally: To be honest, I prefer Matlab's default color scale with the additional orange. IMHO, it makes it easier to correlate the colors to a value.

But all in all, the results are really impressive!

Markus

Reply | Threaded
Open this post in threaded view
|

Re: Request for running compare_plot_demos

Rik-4
On 01/15/2019 10:50 AM, "Markus Mützel" wrote:

> Am Dienstag, 15. Januar 2019 um 13:10 Uhr "Kai Torben Ohlhus" wrote:
>> Anyways, the results look very convincing!  Can we put this on the website after the release maybe? ;-)
>  
> Thanks for running these tests.
> You're right: I also think that these are the best results we had so far.
>
> Just a few observations without looking at any of the demo code:
> - Something seems to be off with the camzoom demo and the qt toolkit.
> - The syntax for our copyobj doesn't seem to be Matlab compatible.
> - Either ours or Matlab's plotboxaspectratio seems to be off. E.g. for demo ellipsoid.
> - We might be doing something wrong with the z axis scale for zoom...
> - And finally: To be honest, I prefer Matlab's default color scale with the additional orange. IMHO, it makes it easier to correlate the colors to a value.
>
> But all in all, the results are really impressive!
It does really look good.  Now, in several cases, we actually look better
than Matlab.  I didn't realize that they also had such a hard time with
subplots, colorbars, and large legends placed outside the axes.  I filed a
bug report to keep track of this:
https://savannah.gnu.org/bugs/index.php?55481.

There were some interesting differences in light objects which need
inspection.  Here is a partial list of things I found.

demo bar 1 : axes ticks are shown on top of the bars
demo text 1 : rotated text is only valid for 0, 90, 180, 270 degrees.  This
is not the case for the FLTK toolkit which also uses gl2ps
demo polar 1 : The axes lines are semi-transparent in Matlab, as they are
for Octave on a rectangular axes
demo copyobj 1 : Completely fails on Matlab, so we must have the wrong syntax
demo errorbar 2 : Completely different plot.
demo ezplot 3 : Complaint from Matlab about undefined variable 'T'
demo annotation 3 : There are more arrowhead styles available for Matlab.
demo box 3 : The 3-D ticklengths are overly large.  This is a problem with
all 3-D plots.
demo ezplot3 1 : Completely different plot
demo ezplot 1 : Matlab replaces special text, like pi, with '\pi' so that
tex interpreter can produce symbols.
demo fplot 1 : Matlab doesn't automatically add a legend, but I actually
like having one.
demo isonormal 1 : Upper left plot is different from Matlab.
demo stem 1 : Matlab draws axes box for stem plots.  Could turn that on
very easily.
demo light 5 : It looks like the reflections in the right column are in the
opposite direction from Matlab (multiplication by -1 maybe?)
demo light 9 : Matlab's "lit" and "unlit" look the same.
demo line 1 : No Matlab plot so we have incorrect syntax somehow.
demo polar 2 : Matlab plot is completely different.
demo polar 4 : polar axes grid lines are under, rather than over, polar plot.
demo scatter 3 : Matlab fails because the color argument C is somehow
incorrect.

I'll be taking care of bar, polar, and stem today.  I'll take a look at the
size of the ticks as well, but that is going to be in C++.

--Rik


Reply | Threaded
Open this post in threaded view
|

Re: Request for running compare_plot_demos

siko1056
In reply to this post by Rik-4
On Tue, Jan 15, 2019 at 6:57 PM Rik <[hidden email]> wrote:
On 01/15/2019 04:10 AM, Kai Torben Ohlhus wrote:
Thus I needed to comment the halve, ran it, commented the other halve, ran it again, and finally merged the diary-outputs.
Additionally, four times the double parenthesis indexing ()() was used, which is not supported by Matlab. I split those occurrences in two statements.

Can you report those four occurrences as a bug?  There's no reason we can't right the %!demo code to run under both Octave and Matlab which would make this easier in the future.

Rik, you just fix problems too fast, or I am just too slow :D

Both mentioned problems with double parenthesis in "plotyy" and "copyobj" you fixed in


Kai
Reply | Threaded
Open this post in threaded view
|

Re: Request for running compare_plot_demos

Rik-4
On 01/16/2019 12:39 AM, Kai Torben Ohlhus wrote:
On Tue, Jan 15, 2019 at 6:57 PM Rik <[hidden email]> wrote:
On 01/15/2019 04:10 AM, Kai Torben Ohlhus wrote:
Thus I needed to comment the halve, ran it, commented the other halve, ran it again, and finally merged the diary-outputs.
Additionally, four times the double parenthesis indexing ()() was used, which is not supported by Matlab. I split those occurrences in two statements.

Can you report those four occurrences as a bug?  There's no reason we can't right the %!demo code to run under both Octave and Matlab which would make this easier in the future.

Rik, you just fix problems too fast, or I am just too slow :D

I am going too fast.  I wrote "we can't right" when I meant "we can't write".  Both words sound the same in English, but are spelled differently and mean very different things.  Anyways, its all cleared up.

--Rik