matplotlib-users Mailing List for matplotlib

On Fri, Jul 29, 2011 at 6:19 PM, Roy Lowrance <roy...@gm...>wrote:

> I'm using matplotlib on ubuntu 11.04.
>
> I create a figure and an axes and then show it via plt.show().
>
> From the window that plt.show() opens, I save the file to plot-3.png. This
> works as I can open the file with evince.
>
> However, when my program executes plt.savefig('plot-3.png'), something is
> saved as a file is created, but when I open the file with evince (or GIMP),
> I see just the canvas, not the figure.
>
> How do I save under program control?
>
> - Roy
>
>
There are a few common mistakes that might have happened.  First, if you did
a plt.show() and then closed the displayed figure, and *then* saved the
figure, you will be saving the image of a new figure as you have already
"destroyed" the old figure.  If this is the case, put the savefig command
*before* plt.show().

Another possibility is that another figure was somehow created between the
last plotting
function and the call to savefig().  plt.savefig() assumes the current
(active)  figure, and so if another figure is accidentially created before
calling savefig(), then you will be saving a blank figure.

I hope this helps, and let us know if you have further questions!
Ben Root

I'm using matplotlib on ubuntu 11.04.

I create a figure and an axes and then show it via plt.show().

>From the window that plt.show() opens, I save the file to plot-3.png. This
works as I can open the file with evince.

However, when my program executes plt.savefig('plot-3.png'), something is
saved as a file is created, but when I open the file with evince (or GIMP),
I see just the canvas, not the figure.

How do I save under program control?

- Roy

Dear users,

I would like to draw lines with fixed pixel thickness, for example 1,
2, or 3 pixels, which remains 1 pixel, even if I zoom in and out the
graph in my interactive application. How can I do this with matplotlib?

I tried to set linewidth to 1 on creating LineCollection, but this
creates too thick lines. If I set it to 0.25 or 0.1, for example, then
some lines on the screen are thin, but some line instances are
simply missing (not visible), depending on the zoom.


Thank you for your help,

János

On Fri, Jul 29, 2011 at 3:12 PM, shooter123 <fr...@gm...> wrote:

>
> Hi,
>
> In the following example -
> http://matplotlib.sourceforge.net/examples/mplot3d/hist3d_demo.html
> how can I set custom tick labels for x and y axes?
> My plot dissappears when using ax.set_xticks
>

Using ax.set_xticks() won't work (at least, not until the upcoming release
where I fixed this).  Try ax.w_xaxis.set_ticks() as a work-around.

Ben Root

Hi,

In the following example -
http://matplotlib.sourceforge.net/examples/mplot3d/hist3d_demo.html
how can I set custom tick labels for x and y axes?
My plot dissappears when using ax.set_xticks
-- 
View this message in context: http://old.nabble.com/How-to-set-custom-tick-labels-for-bar3d-plot-tp32161706p32161706.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

You need to be able to build matplotlib from source (which implies 
having a compiler etc., which may or may not be easy on your platform).  
The basic instructions for building from source are here:

http://matplotlib.sourceforge.net/users/installing.html

and the branch containing the locale fix is available for git checkout 
or download as a tarball here:

https://github.com/mdboom/matplotlib/tree/locale_formatting

Mike

On 07/29/2011 01:14 PM, José Alexandre Nalon wrote:
> Hello,
>
> Thanks for your work! But, since I don't have much experience
> with dealing with the sources and compiling, I don't know
> exactly how to apply this patch. Is there a guide to applying
> patches to matplotlib or something like that? Any help will
> be appreciated.
>
> ---
> José Alexandre Nalon
> na...@te...
>
>
> Em 29/07/2011, às 12:08, Michael Droettboom escreveu:
>
>> There isn't a really good way to do this at present.  Maybe someone can
>> suggest a workaround that doesn't require modifying matplotlib.
>>
>> However, I was able to produce a patch that will respect the user's
>> current locale here:
>>
>> https://github.com/matplotlib/matplotlib/pull/410
>>
>> Cheers,
>> Mike
>>
>> On 07/28/2011 06:46 PM, José Alexandre Nalon wrote:
>>> Hello,
>>>
>>> Sorry if this question was answered before. I did a search in the
>>> mailing list archives but couldn't find any answer.
>>>
>>> Is there an easy way to use comma as a decimal separator in texts,
>>> especially in tick labels? I know that it is possible to user
>>> formatters to do that, but I was wondering if there is an automatic
>>> setting or something that could be applied to axes instances, as I
>>> have more than 200 scripts to generate plots and I don't want to
>>> edit every one of them. Maybe there is a TeX setting (I uset
>>> usetex=True)?
>>>
>>> I tried setting locales, but it didn't work. I also tried to
>>> process the texts in the tick labels, but that didn't work also.
>>> Maybe I am overseeing something, but any hint would help.
>>>
>>> Thanks in advance:
>>>
>>> ---
>>> José Alexandre Nalon
>>> na...@te...
>>>
>>> ------------------------------------------------------------------------------
>>> Got Input?   Slashdot Needs You.
>>> Take our quick survey online.  Come on, we don't ask for help often.
>>> Plus, you'll get a chance to win $100 to spend on ThinkGeek.
>>> http://p.sf.net/sfu/slashdot-survey
>>> _______________________________________________
>>> Matplotlib-users mailing list
>>> Mat...@li...
>>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>> ------------------------------------------------------------------------------
>> Got Input?   Slashdot Needs You.
>> Take our quick survey online.  Come on, we don't ask for help often.
>> Plus, you'll get a chance to win $100 to spend on ThinkGeek.
>> http://p.sf.net/sfu/slashdot-survey
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>
> ------------------------------------------------------------------------------
> Got Input?   Slashdot Needs You.
> Take our quick survey online.  Come on, we don't ask for help often.
> Plus, you'll get a chance to win $100 to spend on ThinkGeek.
> http://p.sf.net/sfu/slashdot-survey
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Hello,

Thanks for your work! But, since I don't have much experience
with dealing with the sources and compiling, I don't know
exactly how to apply this patch. Is there a guide to applying
patches to matplotlib or something like that? Any help will
be appreciated.

---
José Alexandre Nalon
na...@te...


Em 29/07/2011, às 12:08, Michael Droettboom escreveu:

> There isn't a really good way to do this at present.  Maybe someone can 
> suggest a workaround that doesn't require modifying matplotlib.
> 
> However, I was able to produce a patch that will respect the user's 
> current locale here:
> 
> https://github.com/matplotlib/matplotlib/pull/410
> 
> Cheers,
> Mike
> 
> On 07/28/2011 06:46 PM, José Alexandre Nalon wrote:
>> Hello,
>> 
>> Sorry if this question was answered before. I did a search in the
>> mailing list archives but couldn't find any answer.
>> 
>> Is there an easy way to use comma as a decimal separator in texts,
>> especially in tick labels? I know that it is possible to user
>> formatters to do that, but I was wondering if there is an automatic
>> setting or something that could be applied to axes instances, as I
>> have more than 200 scripts to generate plots and I don't want to
>> edit every one of them. Maybe there is a TeX setting (I uset
>> usetex=True)?
>> 
>> I tried setting locales, but it didn't work. I also tried to
>> process the texts in the tick labels, but that didn't work also.
>> Maybe I am overseeing something, but any hint would help.
>> 
>> Thanks in advance:
>> 
>> ---
>> José Alexandre Nalon
>> na...@te...
>> 
>> ------------------------------------------------------------------------------
>> Got Input?   Slashdot Needs You.
>> Take our quick survey online.  Come on, we don't ask for help often.
>> Plus, you'll get a chance to win $100 to spend on ThinkGeek.
>> http://p.sf.net/sfu/slashdot-survey
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
> 
> ------------------------------------------------------------------------------
> Got Input?   Slashdot Needs You.
> Take our quick survey online.  Come on, we don't ask for help often.
> Plus, you'll get a chance to win $100 to spend on ThinkGeek.
> http://p.sf.net/sfu/slashdot-survey
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 

Hello,

I found the following page on running Python using "parallel processing",
but I have no idea which one would be most compatible with matplotlib and
basemap.

http://wiki.python.org/moin/ParallelProcessing

Any help would be greatly appreciated. I looked through the archives and it
looks like in 2009 at the conference there was a session on using mpi4py;
is that still the best option?

Thank you in advance.

Sincerely,

Jeremy Lounds

On Thu, Jul 28, 2011 at 8:44 PM, Joseph Costlow <jos...@gm...>wrote:

> This is a small omission, and I think most people would figure it out, but
> I thought it might be worth mentioning:
>
> Page:
> http://matplotlib.sourceforge.net/users/pyplot_tutorial.html#working-with-multiple-figures-and-axes
>
> The code example, both on the page, and the link is missing:
>
> plt.show()
>
> Best,
>
> Joe
>
>
Actually, all of the examples (on that page) seem to be missing plt.show().
But, in defense of what is currently there, they are also missing the import
calls as well.  Presumedly, one could assume that interactive mode was on,
but this isn't given, either.

I am torn on this one.  On one hand, there is something to be said for
simplicity and focusing only on the relevant part of the code.  On the other
hand, explicit code examples would reduce ambiguity.

What do others think?

Ben Root

There isn't a really good way to do this at present.  Maybe someone can 
suggest a workaround that doesn't require modifying matplotlib.

However, I was able to produce a patch that will respect the user's 
current locale here:

https://github.com/matplotlib/matplotlib/pull/410

Cheers,
Mike

On 07/28/2011 06:46 PM, José Alexandre Nalon wrote:
> Hello,
>
> Sorry if this question was answered before. I did a search in the
> mailing list archives but couldn't find any answer.
>
> Is there an easy way to use comma as a decimal separator in texts,
> especially in tick labels? I know that it is possible to user
> formatters to do that, but I was wondering if there is an automatic
> setting or something that could be applied to axes instances, as I
> have more than 200 scripts to generate plots and I don't want to
> edit every one of them. Maybe there is a TeX setting (I uset
> usetex=True)?
>
> I tried setting locales, but it didn't work. I also tried to
> process the texts in the tick labels, but that didn't work also.
> Maybe I am overseeing something, but any hint would help.
>
> Thanks in advance:
>
> ---
> José Alexandre Nalon
> na...@te...
>
> ------------------------------------------------------------------------------
> Got Input?   Slashdot Needs You.
> Take our quick survey online.  Come on, we don't ask for help often.
> Plus, you'll get a chance to win $100 to spend on ThinkGeek.
> http://p.sf.net/sfu/slashdot-survey
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

This is a small omission, and I think most people would figure it out, but I
thought it might be worth mentioning:

Page:
http://matplotlib.sourceforge.net/users/pyplot_tutorial.html#working-with-multiple-figures-and-axes

The code example, both on the page, and the link is missing:

plt.show()

Best,

Joe