[matplotlib-devel] [Re]: kwarg "range" in hist

[Manuel Metz]

Please see the end of the mail for the important point !!!

Eric Firing wrote:
> Manuel,
> 
> Although it doesn't hurt, I don't think it is worthwhile changing range 
> to xrange.  From the 2.5 docs:
[...snip...]
> Note "minimal" advantage.  xrange was intended for special-case use, not 
> general use.

Eric,

yes, I absolutely agree with you that this is only a small (minimal)
advantage, probably not worth to worry about. Nevertheless ...

> And from Python 3.0, http://docs.python.org/dev/3.0/whatsnew/3.0.html
> xrange() renamed to range(), so range() will no longer produce a list 
> but an iterable yielding integers when iterated over.

Python 3.0 will use xrange() by default, but it is then named range(),
so from that _I_ conclude that xrange() should be used by default. You
can also see the difference by using 2to3:

"""
for i in range(10): print i
for i in xrange(10): print i
"""

gets converted to:

"""
for i in range(10): print i
for i in range(10): print i
"""

That is, because 2to3 is a clever program. But:

"""
a = range(10)
b = xrange(10)
for i in a: print i
for i in b: print i
"""

gets converted to

"""
a = list(range(10))
b = range(10)
for i in a: print(i)
for i in b: print(i)
"""

;-)

As you said, it's only a minimal advantage and 2to3 is a clever code!!!


(THE IMPORTANT POINT)

But this brings me to another, more important point: In the axes hist()
method, a keyword named "range" is used that is passed to the numpy
histogram() function, which has the kwarg 'range'. Now, this is not a
problem as long as the range() builtin function is not used in the
hist() method. But there are a few loops in this method that use
xrange(), so this code will be translated to range() in py3 -- and that
will be a problem. A basic example with a pseudo-code:

"""
def foo(x, range=(1,10)):
print range
for i in xrange(x): print i
foo(10)
"""

with 2to3 -->

"""
def foo(x, range=(1,10)):
print(range)
for i in range(x): print(i)
foo(10)
"""
which then fails.

One solution would be to use a different keyword argument, maybe
"binrange" instead of "range" and to throw a deprecated warning for the
old keyword ???

Manuel

> This implies to me that range is the preferred form, and xrange is 
> intended to go away.
> 
> Eric
> 
[...snip...]


-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel

[matplotlib-devel] creating docs fails

[Manuel Metz]

With a clear checkout building the docs fails:

[...]
Sphinx v0.4.2, building html
trying to load pickled env... not found
building [html]: targets for 348 source files that are out of date
updating environment: 348 added, 0 changed, 0 removed
reading... api/afm_api api/artist_api Exception occurred:
  File "/usr/lib/python2.5/shutil.py", line 51, in copyfile
fsrc = open(src, 'rb')
IOError: [Errno 2] No such file or directory:
'../mpl_examples/pylab_examples/findobj_demo.py'
The full traceback has been saved in /tmp/sphinx-err-X12gbJ.log, if you
want to report the issue to the author.
Please also report this if it was a user error, so that a better error
message can be provided next time.
Send reports to [EMAIL PROTECTED] Thanks!
Building HTML failed.


-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel

[matplotlib-devel] colormap clipping for large numbers (small vmax - vmin)

[Tony S Yu]

I was using pcolor with very large numbers and a small vrange (vmax - vmin), and ran into a float to integer conversion problem. Large numbers get converted to *negative* integers by astype (see numpy thread) in colors.Colormap.__call__.I'm not sure if this is even worth fixing since almost no one would run into this problem (unless you were doing something stupid, like trying to use pcolor as a 2D zero finder :). For the error to occur, you have to set vmin/vmax values (otherwise, the data is properly normalized before converting to integers), and your data has to greatly exceed these limits.Cheers,-Tony#!/usr/bin/env python

from __future__ import with_statement


import numpy as np
import cookbook as cb
vmin = 0.45
vmax = 0.55
N = 1000
with cb.stopwatch():
xa = np.random.sample((N, N))
np.putmask(xa, xa > vmax, vmax)
np.putmask(xa, xa < vmin, vmin)

with cb.stopwatch():
xa = np.random.sample((N, N))
np.clip(xa, vmin, vmax, out=xa)Example of the problem:#~import matplotlib.pyplot as pltimport numpy as npcmap = plt.cm.graycmap.set_over('r', 1.0)cmap.set_under('g', 1.0)cmap.set_bad('b', 1.0)eps = 1E-8A = np.arange(100).reshape(10, 10)plt.pcolor(A, vmin=50, vmax=50+eps, cmap=cmap)# the plot should be about half red and half green (plus a black square)# without patch, some of the red squares are filled greenplt.show()#~-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/___
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel

Re: [matplotlib-devel] colormap clipping for large numbers (small vmax - vmin)

[Tony Yu]

Sorry, I attached the wrong file:

colormap_clipping.patch
Description: Binary data
On Oct 18, 2008, at 11:23 AM, Tony S Yu wrote:I was using pcolor with very large numbers and a small vrange (vmax - vmin), and ran into a float to integer conversion problem. Large numbers get converted to *negative* integers by astype (see numpy thread) in colors.Colormap.__call__.I'm not sure if this is even worth fixing since almost no one would run into this problem (unless you were doing something stupid, like trying to use pcolor as a 2D zero finder :). For the error to occur, you have to set vmin/vmax values (otherwise, the data is properly normalized before converting to integers), and your data has to greatly exceed these limits.Cheers,-TonyExample of the problem:#~import matplotlib.pyplot as pltimport numpy as npcmap = plt.cm.graycmap.set_over('r', 1.0)cmap.set_under('g', 1.0)cmap.set_bad('b', 1.0)eps = 1E-8A = np.arange(100).reshape(10, 10)plt.pcolor(A, vmin=50, vmax=50+eps, cmap=cmap)# the plot should be about half red and half green (plus a black square)# without patch, some of the red squares are filled greenplt.show()#~-This SF.Net email is sponsored by the Moblin Your Move Developer's challengeBuild the coolest Linux based applications with Moblin SDK & win great prizesGrand prize is a trip for two to an Open Source event anywhere in the worldhttp://moblin-contest.org/redirect.php?banner_id=100&url="">Matplotlib-devel mailing listMatplotlib-devel@lists.sourceforge.nethttps://lists.sourceforge.net/lists/listinfo/matplotlib-devel-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/___
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel

Re: [matplotlib-devel] [Re]: kwarg "range" in hist

[Eric Firing]

Manuel Metz wrote:
> Please see the end of the mail for the important point !!!

Thank you--I see you are way ahead of me on this.  See comments below.
> 
> Eric Firing wrote:
>> Manuel,
>>
>> Although it doesn't hurt, I don't think it is worthwhile changing range 
>> to xrange.  From the 2.5 docs:
> [...snip...]
>> Note "minimal" advantage.  xrange was intended for special-case use, not 
>> general use.
> 
> Eric,
> 
> yes, I absolutely agree with you that this is only a small (minimal)
> advantage, probably not worth to worry about. Nevertheless ...
> 
>> And from Python 3.0, http://docs.python.org/dev/3.0/whatsnew/3.0.html
>> xrange() renamed to range(), so range() will no longer produce a list 
>> but an iterable yielding integers when iterated over.
> 
> Python 3.0 will use xrange() by default, but it is then named range(),
> so from that _I_ conclude that xrange() should be used by default. You
> can also see the difference by using 2to3:
> 
> """
> for i in range(10): print i
> for i in xrange(10): print i
> """
> 
> gets converted to:
> 
> """
> for i in range(10): print i
> for i in range(10): print i
> """
> 
> That is, because 2to3 is a clever program. But:
> 
> """
> a = range(10)
> b = xrange(10)
> for i in a: print i
> for i in b: print i
> """
> 
> gets converted to
> 
> """
> a = list(range(10))
> b = range(10)
> for i in a: print(i)
> for i in b: print(i)
> """
> 
> ;-)
> 
> As you said, it's only a minimal advantage and 2to3 is a clever code!!!
> 

I am glad you brought the above to my attention--it completely changes 
my point of view.  It does appear that changing to xrange now, whenever 
it will work (that is, when one does not *need* a list) will make the 
transition to Python 3 more efficient and has no disadvantage with 
present code.

> 
> (THE IMPORTANT POINT)
> 
> But this brings me to another, more important point: In the axes hist()
> method, a keyword named "range" is used that is passed to the numpy
> histogram() function, which has the kwarg 'range'. Now, this is not a
> problem as long as the range() builtin function is not used in the
> hist() method. But there are a few loops in this method that use
> xrange(), so this code will be translated to range() in py3 -- and that
> will be a problem. A basic example with a pseudo-code:
> 
> """
> def foo(x, range=(1,10)):
> print range
> for i in xrange(x): print i
> foo(10)
> """
> 
> with 2to3 -->
> 
> """
> def foo(x, range=(1,10)):
> print(range)
> for i in range(x): print(i)
> foo(10)
> """
> which then fails.
> 
> One solution would be to use a different keyword argument, maybe
> "binrange" instead of "range" and to throw a deprecated warning for the
> old keyword ???
> 

Yes, I think the use of any builtin as a kwarg is a bug that should be 
squashed via a new kwarg with a deprecation.  Similarly, use of any 
builtin as in internal variable should be considered a latent bug and fixed.

Unfortunately, in this case, the badly-named kwarg is in numpy.histogram 
as well.  The best thing would be to try to get the same change made in 
numpy so that mpl hist and numpy.histogram kwargs would stay in sync.

To make matters worse, histogram has evolved in such a way that its 
kwargs are a confusing mess.  It is too bad that when the "new" syntax 
was developed, the "range" kwarg was not changed at the same time. I 
don't know whether any more changes would be accepted now.

If there is to be a new kwarg, I think I would call it "cliprange", 
since it is essentially used to clip the input--unless "new" is not 
True.  It is not really a "bin range", because it can be set 
independently of the bins.  (I have not traced the code; I am basing my 
statement on the docstring, so I could be wrong about what the code 
actually does.)

Eric

> Manuel
> 
>> This implies to me that range is the preferred form, and xrange is 
>> intended to go away.
>>
>> Eric
>>
> [...snip...]
> 
> 
> -
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> ___
> Matplotlib-devel mailing list
> Matplotlib-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel


-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourcef

Re: [matplotlib-devel] colormap clipping for large numbers (small vmax - vmin)

[Eric Firing]

Tony S Yu wrote:
> I was using pcolor with very large numbers and a small vrange (vmax - 
> vmin), and ran into a float to integer conversion problem. Large numbers 
> get converted to *negative* integers by astype (see numpy thread 
> )
>  in 
> colors.Colormap.__call__.
> 
> I'm not sure if this is even worth fixing since almost no one would run 
> into this problem (unless you were doing something stupid, like trying 
> to use pcolor as a 2D zero finder :). For the error to occur, you have 
> to set vmin/vmax values (otherwise, the data is properly normalized 
> before converting to integers), and your data has to greatly exceed 
> these limits.

Tony,

Thank you.

I committed the change; it looks like the cost of the extra clip is 
negligible, and it is nice to make the behavior correct even under 
extreme conditions.

Eric

> 
> Cheers,
> -Tony
> 
> 
> 
> 
> 
> Example of the problem:
> #~
> import matplotlib.pyplot as plt
> import numpy as np
> 
> cmap = plt.cm.gray
> cmap.set_over('r', 1.0)
> cmap.set_under('g', 1.0)
> cmap.set_bad('b', 1.0)
> 
> eps = 1E-8
> 
> A = np.arange(100).reshape(10, 10)
> plt.pcolor(A, vmin=50, vmax=50+eps, cmap=cmap)
> # the plot should be about half red and half green (plus a black square)
> # without patch, some of the red squares are filled green
> plt.show()
> #~
> 
> 
> 
> 
> -
> This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
> Build the coolest Linux based applications with Moblin SDK & win great prizes
> Grand prize is a trip for two to an Open Source event anywhere in the world
> http://moblin-contest.org/redirect.php?banner_id=100&url=/
> 
> 
> 
> 
> ___
> Matplotlib-devel mailing list
> Matplotlib-devel@lists.sourceforge.net
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel


-
This SF.Net email is sponsored by the Moblin Your Move Developer's challenge
Build the coolest Linux based applications with Moblin SDK & win great prizes
Grand prize is a trip for two to an Open Source event anywhere in the world
http://moblin-contest.org/redirect.php?banner_id=100&url=/
___
Matplotlib-devel mailing list
Matplotlib-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/matplotlib-devel

Re: [matplotlib-devel] [Fwd: Re: Patch for scatter plot legend enhancement]

[Jae-Joon Lee]

> To help clarify the original purpose of  "update_from": I wrote this
> method when writing the original legend implementation so the legend
> proxy objects could easily copy their style attributes from the
> underlying objects they were a proxy for (so not every property is
> copied, eg the xdata for line objects is not copied).  So the
> operating question should be: what properties do I need to copy to
> make the legend representation of the object.   While you are in
> there, perhaps you could clarify this in the docstrings of the
> update_from method.

Thanks for clarifying this, John.

Manuel,
The patch looks good to me. We may submit the patch (I hope Erik is
okay with the current patch) and it would be great if you handle the
submission.

-JJ





On Fri, Oct 17, 2008 at 9:45 PM, Manuel Metz <[EMAIL PROTECTED]> wrote:
> Jae-Joon Lee wrote:
>> Thanks Manuel.
>>
>> Yes, we need rotation value and etc, but my point is, do we need to
>> update it within the update_from() method? Although my preference is
>> not to do it, it may not matter much as far as we state what this
>> method does clearly in the doc.
>
> Okay, it's probably better to create the object correctly (numsides ...)
> instead of copying the properties (see also JDHs mail !)
>
>> And, in your patch, I don't think updating the numsides value has any
>> effect as it does not recreate the paths.
>>
>> I'm attaching the revised patch. In this patch, update_from() only
>> update gc-related properties. And numsides, size, and rotations are
>> given during the object creation time.
>
> Yes, this looks better. But creating handle_sizes is a little bit too
> much effort. This is done internally. It will do passing a sizes list,
> that may or may not be shorter/longer than numpoints (see revised patch).
>
> I also changed the way the yoffsets are updated in _update_positions().
>
> One additional thing I have in mind (for a later time) is a "sizesbar"
> similar to a colorbar where you can read off values corresponding to
> marker sizes...
>
> Cheers,
>  Manuel
>
>> Erik,
>> I see your points. My main concern is that the yoffsets makes the
>> results a bit funny when numpoints is 2. The attached patch has a
>> varying sizes of [0.5*(max+min), max, min]. The yoffsets are only
>> introduced when numpints > 2 and you can also provide it as an
>> optional argument.
>>
>> Regards,
>>
>> -JJ
>>
>>
>> On Thu, Oct 16, 2008 at 8:43 PM, Manuel Metz <[EMAIL PROTECTED]> wrote:
>>> Manuel Metz wrote:
 Jae-Joon Lee wrote:
> Hi Manuel,
>
> I think it is a good to introduce the update_from method in Collections.
> But, I'm not sure if it is a good idea to also update sizes, paths and
> rotation (in RegularPolyCoolection). My impression is that update_from
> method is to update gc related attributes. For comparison,
> Patch.update_from() does not update the path.
 That's exactly the point why I wasn't fully happy with the patch. The
 path is generated by the _path_generator, so instead of copying the path
  it seems to be better to create an instance of the corresponding class
 (e.g. the StarPolygonCollection class, as suggested before).

   One should update the rotation attribute (!!); it's only one number. A
 '+' marker, for example, has rotation = 0, whereas a 'x' marker has
 rotation=pi/4. That's the only difference between those two !

> Also, is it okay to update properties without checking its length?. It
> does not seem to cause any problems though.
   It's in principal not a problem to copy the sizes attribute without
 checking the length. If it's shorter the the number of items the sizes
 are repeated; if it's longer it gets truncated.

 mm

> I guess It would better to use xdata_markers than xdata in the
> get_handle() method. The difference is when numpoints==1. Using xdata
> gives two marker points.
>
> I was actually about to to commit my patch. I'll try to account your
> changes and post my version of patch later today.
>
> Regards,
>
> -JJ
>
>
>
>
> On Wed, Oct 15, 2008 at 4:07 PM, Manuel Metz <[EMAIL PROTECTED]> wrote:
>> hmm
>>
>>  Original Message 
>> Jae-Joon Lee wrote:
 - the parameter numpoints should be used (it's ignored right now)

>>> Thanks Manuel. I guess we can simply reuse xdata_marker for this 
>>> purpose.
>>>
>>>
 - Some private variables are accessed and a new RegularPolycollection 
 is
 created (does this work eg. with a StarPolygonCollection? I haven't
 checked, but I don't think so !). Instead of creating a new
 RegularPolyCollection it might be more useful to make a copy of the
 existing object... I was thinking about a update_from() method for the
 Collection class(es) similar to update_from() for lines.

>>> By changing "RegularPolyCoole