matplotlib-users Mailing List for matplotlib

Rory,

I applied your patch to svn, revision 2408.

Thanks.

Eric

Rory Yorke wrote:
> Dear list,
> 
> There is a bug in stem, demonstrated by:
> 
> 
> from matplotlib.pylab import *
> 
> stem(arange(10),arange(10),hold=False)
> show()
> 
> 
> A simple fix is below; it simply sets hold to True for the duration of
> the stem method, and restores the hold state once done.
> 
> This was tested with the very simple
> 
> 
> plot(arange(10)); stem(arange(10),arange(10),hold=True); show()
> 
> 
> and
> 
> 
> plot(arange(10)); stem(arange(10),arange(10),hold=False); show()
> 
> 
> Regards,
> 
> Rory
> 
> Index: axes.py
> ===================================================================
> --- axes.py	(revision 2406)
> +++ axes.py	(working copy)
> @@ -2509,7 +2509,10 @@
>          http://www.mathworks.com/access/helpdesk/help/techdoc/ref/stem.html
>          for details and examples/stem_plot.py for a demo.
>          """
> +        remember_hold=self._hold
>          if not self._hold: self.cla()
> +        self.hold(True)
> +
>          markerline, = self.plot(x, y, markerfmt)
>  
>          stemlines = []
> @@ -2518,6 +2521,9 @@
>              stemlines.append(l)
>  
>          baseline, = self.plot([amin(x), amax(x)], [0,0], basefmt)
> +
> +        self.hold(remember_hold)
> +
>          return markerline, stemlines, baseline
>  
>  
> 
> 
> 
> 
> -------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job easier
> Download IBM WebSphere Application Server v.1.0.1 based on Apache Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

Try passing the "loc" kwarg, like loc=(1.05, 0) to the legend command.

plot([1,2])
legend(['line1'], loc=(1.05,0))

Darren


On Saturday 20 May 2006 13:48, Ryan Krauss wrote:
> Is there a built-in way for pylab to put the legend off to the right
> of the plot.  I need to have a fairly long legend (12 entries) and I
> need to figure out what to do with it.  (The plot is not as ugly as it
> sounds with 12 things on it.)
>
> Ryan
>
>
> -------------------------------------------------------
> Using Tomcat but need to do more? Need to support web services, security?
> Get stuff done quickly with pre-integrated technology to make your job
> easier Download IBM WebSphere Application Server v.1.0.1 based on Apache
> Geronimo
> http://sel.as-us.falkag.net/sel?cmd=lnk&kid=120709&bid=263057&dat=121642
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

-- 
Darren S. Dale, Ph.D.
Cornell High Energy Synchrotron Source
Cornell University
200L Wilson Lab
Rt. 366 & Pine Tree Road
Ithaca, NY 14853

dd...@co...
office: (607) 255-9894
fax: (607) 255-9001

Is there a built-in way for pylab to put the legend off to the right
of the plot.  I need to have a fairly long legend (12 entries) and I
need to figure out what to do with it.  (The plot is not as ugly as it
sounds with 12 things on it.)

Ryan

>>>>> "Ryan" == Ryan Krauss <rya...@gm...> writes:

    Ryan> I have to plot some things that are the results of
    Ryan> hyperbolic sines and cosines.  This leads to large
    Ryan> amplitudes on the y-axis.  I am having trouble saving these
    Ryan> plots to eps files.  But I have done this successfully
    Ryan> before.

    Ryan> These lines

    Ryan> b=arange(1,40,0.5) y=sinh(b) figure(5) plot(b,y)
    Ryan> savefig('test3.eps') savefig('test3.png')

    Ryan> where the png is fine and the eps shows only an empty axis.

    Ryan> I have to get my thesis to my committee next week and I am
    Ryan> freaking out a bit.

It looks fine for me using matplotlib svn -- try dropping the attached
backend_ps file into your site-packages/matplotlib/backends/ dir

Darren went through some last minute matplotlib thesis panics too,
mainly over mathtext, but got through.  Breath deeply -- I'm sure we
can get it worked out....

On Saturday 20 May 2006 13:03, Ryan Krauss wrote:
> Updating to the latest svn seems to have solved my problems.

Thats good to hear. Your problem may have been caused by the conversion of the 
ps backend to use the new api, which may have been incomplete at the time of 
your last update. I believe the backend has stabilized.

> One weird note though, it seems like the version I was using (svn from
> a month or so ago) was more tolerant to latex errors.  The first time
> after I updated, I got errors on some of the scripts I wrote
> yesterday.  The current version gives a message about Latex not being
> able to process my file.  The old version made it through.  It turns
> out that I has omitted open and/or closing $'s in the Latex x and y
> labels.  Some how the old version handled '\\beta' and '$\\omega'
> just fine.  It is probably better to just print out the Latex errors
> and not try to guess how to fix it, but the older version was guessing
> correctly.

I put some effort into making usetex behave this way. What you are seeing is 
latex returning an exit status indicating an error. When that happens, mpl 
raises an error and passes along the latex error. I think this is much better 
than failing silently and hoping for the best, because often latex would not 
guess correctly (\frac{1}{2} vs $\frac{1}{2}$ is a good example). 

I apologize for any trouble it may cause you, especially considering you are 
finishing up your thesis. But this change was necessary in order to help me 
track down problems, which was becoming time consuming.

Darren

Updating to the latest svn seems to have solved my problems.

One weird note though, it seems like the version I was using (svn from
a month or so ago) was more tolerant to latex errors.  The first time
after I updated, I got errors on some of the scripts I wrote
yesterday.  The current version gives a message about Latex not being
able to process my file.  The old version made it through.  It turns
out that I has omitted open and/or closing $'s in the Latex x and y
labels.  Some how the old version handled '\\beta' and '$\\omega'
just fine.  It is probably better to just print out the Latex errors
and not try to guess how to fix it, but the older version was guessing
correctly.

FWIH,

Ryan

On 5/20/06, Darren Dale <dd...@co...> wrote:
> On Saturday 20 May 2006 12:23, you wrote:
> > I am installing the latest svn version right now.  We will see what
> > happens.
>
> I understand your caution. Dealing with thesis deadlines can be stressful.
>
> I tried your example with svn mpl, with usetex and without, and the eps output
> looked fine.
>
> Darren
>
> > On 5/20/06, Ryan Krauss <rya...@gm...> wrote:
> > > '0.88svn'
> > >
> > > I am a little nervous about changing versions and having something
> > > break at this point, but if that is what it takes.
> > >
> > > Ryan
> > >
> > > On 5/20/06, Darren Dale <dd...@co...> wrote:
> > > > On Saturday 20 May 2006 11:36 am, you wrote:
> > > > > I have to plot some things that are the results of hyperbolic sines
> > > > > and cosines.  This leads to large amplitudes on the y-axis.  I am
> > > > > having trouble saving these plots to eps files.  But I have done this
> > > > > successfully before.
> > > > >
> > > > > These lines
> > > > >
> > > > > b=arange(1,40,0.5)
> > > > > y=sinh(b)
> > > > > figure(5)
> > > > > plot(b,y)
> > > > > savefig('test3.eps')
> > > > > savefig('test3.png')
> > > > >
> > > > > where the png is fine and the eps shows only an empty axis.
> > > > >
> > > > > I have to get my thesis to my committee next week and I am freaking
> > > > > out a bit.
> > > > >
> > > > > Please help me.
> > > >
> > > > What version are you using
>
> --
> Darren S. Dale, Ph.D.
> Cornell High Energy Synchrotron Source
> Cornell University
> 200L Wilson Lab
> Rt. 366 & Pine Tree Road
> Ithaca, NY 14853
>
> dd...@co...
> office: (607) 255-9894
> fax: (607) 255-9001
>

I have to plot some things that are the results of hyperbolic sines
and cosines.  This leads to large amplitudes on the y-axis.  I am
having trouble saving these plots to eps files.  But I have done this
successfully before.

These lines

b=arange(1,40,0.5)
y=sinh(b)
figure(5)
plot(b,y)
savefig('test3.eps')
savefig('test3.png')

where the png is fine and the eps shows only an empty axis.

I have to get my thesis to my committee next week and I am freaking out a bit.

Please help me.

Thanks,

Ryan

Dear list,

There is a bug in stem, demonstrated by:


from matplotlib.pylab import *

stem(arange(10),arange(10),hold=False)
show()


A simple fix is below; it simply sets hold to True for the duration of
the stem method, and restores the hold state once done.

This was tested with the very simple


plot(arange(10)); stem(arange(10),arange(10),hold=True); show()


and


plot(arange(10)); stem(arange(10),arange(10),hold=False); show()


Regards,

Rory

Index: axes.py
===================================================================
--- axes.py	(revision 2406)
+++ axes.py	(working copy)
@@ -2509,7 +2509,10 @@
         http://www.mathworks.com/access/helpdesk/help/techdoc/ref/stem.html
         for details and examples/stem_plot.py for a demo.
         """
+        remember_hold=self._hold
         if not self._hold: self.cla()
+        self.hold(True)
+
         markerline, = self.plot(x, y, markerfmt)
 
         stemlines = []
@@ -2518,6 +2521,9 @@
             stemlines.append(l)
 
         baseline, = self.plot([amin(x), amax(x)], [0,0], basefmt)
+
+        self.hold(remember_hold)
+
         return markerline, stemlines, baseline