matplotlib-devel Mailing List for matplotlib (Page 3)

On Fri, Sep 23, 2011 at 9:59 AM, Benjamin Root <ben...@ou...> wrote:
>
>
> On Friday, September 23, 2011, John Hunter <jd...@gm...> wrote:
> > Today is the "go-live" date for the mpl release candidate. There are
> > no more open issues tagged release_critical, so I suggest we branch
> > this afternoon. This will give people a last chance to merge in
> > remaining pull requests. If you know you are planning to handle a
> > given pull request before the branch, please drop a comment on the
> > pull request or respond here so we don't duplicate effort. Michael,
> > are you available to make the release branch this afternoon?
> >
>
> There is a bug report (don't have the number on me) that reported a linked
> problem on mac osx on Lion. Has anyone confirmed it?
>
>
Whoops, it wasn't a linker error: Issue #471 "Building macosx backend
ignores arch flags"
Is this going to be a problem for the release?
Ben Root

On Fri, Sep 23, 2011 at 11:56 AM, Michael Droettboom <md...@st...> wrote:
> I agree with your assessment -- it seems like a low-level difference in the
> freetype renderer. I guess we should increase the tolerance on these tests
> (by passing a tol keyword to the image_comparison decorator in
> test_mathtext.py). The default is 1e-3 -- we just need to find the minimum
> value that makes these pass on your machine.
OK, I'll play with this for a bit and see if I can find a better value.
JDH

I agree with your assessment -- it seems like a low-level difference in 
the freetype renderer. I guess we should increase the tolerance on 
these tests (by passing a tol keyword to the image_comparison decorator 
in test_mathtext.py). The default is 1e-3 -- we just need to find the 
minimum value that makes these pass on your machine.
Mike
On 09/23/2011 12:37 PM, John Hunter wrote:
> On Fri, Sep 23, 2011 at 10:34 AM, Michael Droettboom<md...@st...> wrote:
>> I'm not able to reproduce this here (git/master on Ubuntu 10.4, Python
>> 2.7). Can you send and/or link to an example broken image? The way in
>> which it is failing may illustrate the cause of the problem. It could
>> also be that the difference in the freetype library/settings is causing
>> enough difference to trip the threshold (we've seen that before). If
>> that's the case (i.e. they look identical to the human eye, but there
>> are some low-level differences in the anti-aliasing), we probably just
>> need to increase the difference threshold on those tests. If, however,
>> those tests worked on this system until recently, "git bisect" may
>> reveal what caused the breakage.
>
> The images are extremely close by eye: I can see a little difference
> on the font weight on the delta subscript in the attached images if I
> toggle back and forth in eog, but it does look like a tolerance issue.
> I don't exactly when this stopped working, but my platform info is:
>
> BUILDING MATPLOTLIB
> matplotlib: 1.1.0
> python: 2.6.6 (r266:84292, Sep 15 2010, 16:22:56) [GCC
> 4.4.5]
> platform: linux2
>
> REQUIRED DEPENDENCIES
> numpy: 2.0.0.dev-aded70c
> freetype2: 12.0.6
>
> OPTIONAL BACKEND DEPENDENCIES
> libpng: 1.2.44
> Tkinter: Tkinter: 73770, Tk: 8.5, Tcl: 8.5
> Gtk-Message: Failed to load module "gnomesegvhandler":
> libgnomesegvhandler.so: cannot open shared object file: No such file
> or directory
> Gtk-Message: Failed to load module "gnomesegvhandler":
> libgnomesegvhandler.so: cannot open shared object file: No such file
> or directory
> Gtk+: gtk+: 2.22.0, glib: 2.26.1, pygtk: 2.21.0,
> pygobject: 2.21.5
> Mac OS X native: no
> Qt: Qt: 3.3.8, PyQt: 3.18.1
> Qt4: Qt: 4.7.0, PyQt4: 4.7.4
> Cairo: 1.8.8
>
> OPTIONAL DATE/TIMEZONE DEPENDENCIES
> datetime: present, version unknown
> dateutil: 1.5
> /home/jdhunter/dev/lib/python2.6/site-packages/pytz-2010h-py2.6.egg/pytz/__init__.py:32:
> UserWarning: Module dateutil was already imported from
> /home/jdhunter/dev/lib/python2.6/site-packages/dateutil/__init__.pyc,
> but /usr/lib/pymodules/python2.6 is being added to sys.path
> pytz: 2010h
>
> OPTIONAL USETEX DEPENDENCIES
> dvipng: 1.13
> ghostscript: 8.71
> latex: 3.1415926
> pdftops: 0.14.3

On Fri, Sep 23, 2011 at 10:34 AM, Michael Droettboom <md...@st...> wrote:
> I'm not able to reproduce this here (git/master on Ubuntu 10.4, Python
> 2.7). Can you send and/or link to an example broken image? The way in
> which it is failing may illustrate the cause of the problem. It could
> also be that the difference in the freetype library/settings is causing
> enough difference to trip the threshold (we've seen that before). If
> that's the case (i.e. they look identical to the human eye, but there
> are some low-level differences in the anti-aliasing), we probably just
> need to increase the difference threshold on those tests. If, however,
> those tests worked on this system until recently, "git bisect" may
> reveal what caused the breakage.
The images are extremely close by eye: I can see a little difference
on the font weight on the delta subscript in the attached images if I
toggle back and forth in eog, but it does look like a tolerance issue.
 I don't exactly when this stopped working, but my platform info is:
BUILDING MATPLOTLIB
 matplotlib: 1.1.0
 python: 2.6.6 (r266:84292, Sep 15 2010, 16:22:56) [GCC
 4.4.5]
 platform: linux2
REQUIRED DEPENDENCIES
 numpy: 2.0.0.dev-aded70c
 freetype2: 12.0.6
OPTIONAL BACKEND DEPENDENCIES
 libpng: 1.2.44
 Tkinter: Tkinter: 73770, Tk: 8.5, Tcl: 8.5
Gtk-Message: Failed to load module "gnomesegvhandler":
libgnomesegvhandler.so: cannot open shared object file: No such file
or directory
Gtk-Message: Failed to load module "gnomesegvhandler":
libgnomesegvhandler.so: cannot open shared object file: No such file
or directory
 Gtk+: gtk+: 2.22.0, glib: 2.26.1, pygtk: 2.21.0,
 pygobject: 2.21.5
 Mac OS X native: no
 Qt: Qt: 3.3.8, PyQt: 3.18.1
 Qt4: Qt: 4.7.0, PyQt4: 4.7.4
 Cairo: 1.8.8
OPTIONAL DATE/TIMEZONE DEPENDENCIES
 datetime: present, version unknown
 dateutil: 1.5
/home/jdhunter/dev/lib/python2.6/site-packages/pytz-2010h-py2.6.egg/pytz/__init__.py:32:
UserWarning: Module dateutil was already imported from
/home/jdhunter/dev/lib/python2.6/site-packages/dateutil/__init__.pyc,
but /usr/lib/pymodules/python2.6 is being added to sys.path
 pytz: 2010h
OPTIONAL USETEX DEPENDENCIES
 dvipng: 1.13
 ghostscript: 8.71
 latex: 3.1415926
 pdftops: 0.14.3

I'm not able to reproduce this here (git/master on Ubuntu 10.4, Python 
2.7). Can you send and/or link to an example broken image? The way in 
which it is failing may illustrate the cause of the problem. It could 
also be that the difference in the freetype library/settings is causing 
enough difference to trip the threshold (we've seen that before). If 
that's the case (i.e. they look identical to the human eye, but there 
are some low-level differences in the anti-aliasing), we probably just 
need to increase the difference threshold on those tests. If, however, 
those tests worked on this system until recently, "git bisect" may 
reveal what caused the breakage.
Mike
On 09/23/2011 10:59 AM, John Hunter wrote:
> On Fri, Sep 23, 2011 at 8:02 AM, John Hunter<jd...@gm...> wrote:
>> Today is the "go-live" date for the mpl release candidate. There are
>> no more open issues tagged release_critical, so I suggest we branch
>> this afternoon. This will give people a last chance to merge in
>> remaining pull requests. If you know you are planning to handle a
>> given pull request before the branch, please drop a comment on the
>> pull request or respond here so we don't duplicate effort. Michael,
>> are you available to make the release branch this afternoon?
>
> I'm getting tons of failures like the one below relating to mathtext
> fonts on two different platforms (linux and solaris). Anyone else
> seeing these? I did flush my fonts and tex cache and did a clean
> install.
>
>
> ======================================================================
> FAIL: matplotlib.tests.test_mathtext.mathfont_stix_49_test.test
> ----------------------------------------------------------------------
> Traceback (most recent call last):
> File "/usr/lib/pymodules/python2.6/nose/case.py", line 183, in runTest
> self.test(*self.arg)
> File "/home/jdhunter/dev/lib/python2.6/site-packages/matplotlib/testing/decorators.py",
> line 35, in failer
> result = f(*args, **kwargs)
> File "/home/jdhunter/dev/lib/python2.6/site-packages/matplotlib/testing/decorators.py",
> line 126, in do_test
> '(RMS %(rms).3f)'%err)
> ImageComparisonFailure: images not close:
> /home/jdhunter/tmp/mpl_test/result_images/test_mathtext/mathfont_stix_49.png
> vs. /home/jdhunter/tmp/mpl_test/result_images/test_mathtext/expected-mathfont_stix_49.png
> (RMS 11.269)
>
> ------------------------------------------------------------------------------
> All of the data generated in your IT infrastructure is seriously valuable.
> Why? It contains a definitive record of application performance, security
> threats, fraudulent activity, and more. Splunk takes this data and makes
> sense of it. IT sense. And common sense.
> http://p.sf.net/sfu/splunk-d2dcopy2
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel

On Fri, Sep 23, 2011 at 8:02 AM, John Hunter <jd...@gm...> wrote:
> Today is the "go-live" date for the mpl release candidate. There are
> no more open issues tagged release_critical, so I suggest we branch
> this afternoon. This will give people a last chance to merge in
> remaining pull requests. If you know you are planning to handle a
> given pull request before the branch, please drop a comment on the
> pull request or respond here so we don't duplicate effort. Michael,
> are you available to make the release branch this afternoon?
I'm getting tons of failures like the one below relating to mathtext
fonts on two different platforms (linux and solaris). Anyone else
seeing these? I did flush my fonts and tex cache and did a clean
install.
======================================================================
FAIL: matplotlib.tests.test_mathtext.mathfont_stix_49_test.test
----------------------------------------------------------------------
Traceback (most recent call last):
 File "/usr/lib/pymodules/python2.6/nose/case.py", line 183, in runTest
 self.test(*self.arg)
 File "/home/jdhunter/dev/lib/python2.6/site-packages/matplotlib/testing/decorators.py",
line 35, in failer
 result = f(*args, **kwargs)
 File "/home/jdhunter/dev/lib/python2.6/site-packages/matplotlib/testing/decorators.py",
line 126, in do_test
 '(RMS %(rms).3f)'%err)
ImageComparisonFailure: images not close:
/home/jdhunter/tmp/mpl_test/result_images/test_mathtext/mathfont_stix_49.png
vs. /home/jdhunter/tmp/mpl_test/result_images/test_mathtext/expected-mathfont_stix_49.png
(RMS 11.269)

On Friday, September 23, 2011, John Hunter <jd...@gm...> wrote:
> Today is the "go-live" date for the mpl release candidate. There are
> no more open issues tagged release_critical, so I suggest we branch
> this afternoon. This will give people a last chance to merge in
> remaining pull requests. If you know you are planning to handle a
> given pull request before the branch, please drop a comment on the
> pull request or respond here so we don't duplicate effort. Michael,
> are you available to make the release branch this afternoon?
>
There is a bug report (don't have the number on me) that reported a linked
problem on mac osx on Lion. Has anyone confirmed it?
P.S. - l will be finishing up doc work. Let me know if anything needs to be
added to the "what's new" section.
Ben Root

On Fri, Sep 23, 2011 at 9:11 AM, Jouni K. Seppänen <jk...@ik...> wrote:
> This pull request fixes a test failure:
>
> https://github.com/matplotlib/matplotlib/pull/476
>
> Should I just merge it myself, or does someone still want to review it?
Since you and Mike have both looked at it, it is probably good to go.

John Hunter <jd...@gm...> writes:
> Today is the "go-live" date for the mpl release candidate. There are
> no more open issues tagged release_critical, so I suggest we branch
> this afternoon. This will give people a last chance to merge in
> remaining pull requests.
This pull request fixes a test failure:
https://github.com/matplotlib/matplotlib/pull/476
Should I just merge it myself, or does someone still want to review it?
-- 
Jouni K. Seppänen
http://www.iki.fi/jks

Today is the "go-live" date for the mpl release candidate. There are
no more open issues tagged release_critical, so I suggest we branch
this afternoon. This will give people a last chance to merge in
remaining pull requests. If you know you are planning to handle a
given pull request before the branch, please drop a comment on the
pull request or respond here so we don't duplicate effort. Michael,
are you available to make the release branch this afternoon?

On Thu, Sep 22, 2011 at 7:00 PM, Benjamin Root <ben...@ou...> wrote:
> On Thursday, September 22, 2011, Tony Yu <ts...@gm...> wrote:
>> On Thu, Sep 22, 2011 at 5:16 PM, Nathaniel Smith <nj...@po...> wrote:
>>> I looked at the paper, and the goal was specifically to produce a good
>>> "default" colormap - not necessarily the best for any situation, but good
>>> overall and certainly better than the rainbow ('jet') colormap in most
>>> cases. (I agree with the author that jet is pretty terrible and tends to
>>> distort data.)
>>>
>>> Should we switch to this as the default matplotlib colormap? I think it
>>> would be a clear improvement.
>>
>> I have absolutely no clout here, but I'd definitely be in favor of
>> changing the default colormap away from "jet".
>>
>> Personally, I'd prefer a two-tone colormap as the default (two-distinct
>> tones at the limits with a gradient in-between---dubbed "sequential" in the
>> paper) instead of a three-tone colormap (three-distinct tones---dubbed
>> "diverging" in the paper). (I think this is a more common use case, and I
>> think using a "diverging" colormap effectively requires setting vmin/vmax.)
>> But really, (almost) anything is better than "jet".
For those following along, the article is here:
http://www.cs.unm.edu/~kmorel/documents/ColorMaps/ColorMapsExpanded.pdf
The discussion about whether to use a "sequential" or "diverging" map
is in section 3.
I had the same gut reaction as you, but found the paper fairly
convincing. I'm used to diverging maps that really highlight the
center point, like matplotlib's RdBu colormap, and use them all the
time for data where the zero point really matters (and set vmin/vmax
appropriately, like you say). But this colormap is actually really
different from the ones I'm used to, because it's really designed
*not* to highlight the center point as being special. The clearest
example of this is Figure 8 in the paper -- the image on the left is
the one that you'd get from something like RdBu, and the one on the
right is what this colormap produces. And, like he says, it gives you
better detail than a sequential map could.
I actually *wouldn't* want to use this for images that I currently use
RdBu for. But I'm picky -- the map he suggests would still make a heck
of a better colormap for those images than jet does, and Rdbu *really*
isn't appropriate for general use.
On Thu, Sep 22, 2011 at 7:00 PM, Benjamin Root <ben...@ou...> wrote:
> Anyway, this is certainly is worthy of debate, but it certainly won't happen
> for this release. We should be cutting RC tomorrow.
>
> After the release, I encourage you guys to make your cases. Show us plots
> that have been in "jet" and show them as better in another colormap.
Figure 16 in that paper (page 17) is a good start. In the examples
given, for both of the top two jet actively distorts the features of
the data. In row 1, it makes it look like the bars taper off as you
move towards the top of the image, which they don't (compare to the
greyscale image for reference). In row 2, it makes it look like the
"circles" vary in size across the image (which, again, they don't). In
the bottom two images, jet introduces apparent asymmetries that aren't
there: in row 3, you have these 5 apparent stripes of unequal width,
and the yellow is narrower than the cyan. In row 4, well, it's just
obviously terribly misleading. (This isn't surprising, since 'jet'
sweeps through the frequencies of visible light; the big band of
yellow in simulated deuteranope vision corresponds to where they're
missing one of the spectral sensors that most of us have.)
If you want more examples though I can certainly pull some out of my thesis :-).
> As a bit of a challenge to you all, I am not color-blind, but I do wear
> tinted glasses that make it difficult to tell the difference between darker
> blues and black, and sometimes greens and blues are hard to distinguish.
> Furthermore, as a radar meteorologist, I am very accustomed to the
> colormaps commonly used for radar reflectivies (and is similar to "jet").
No colormap is going to be perfect for everyone, and maybe someone
else will pop up with a pointer to a map that's even better supported
than this one. I just think that jet has sufficiently manifest flaws
that it would be a great favor to science if we could switch to
*something* better as our default.
-- Nathaniel

On Thursday, September 22, 2011, Tony Yu <ts...@gm...> wrote:
> On Thu, Sep 22, 2011 at 5:16 PM, Nathaniel Smith <nj...@po...> wrote:
>>
>> On Sep 21, 2011 5:29 PM, "Christoph Gohlke" <cg...@uc...> wrote:
>> > On 9/13/2011 12:24 AM, Eric Firing wrote:
>> > > On 07/18/2011 07:07 AM, Sameer Grover wrote:
>> > >> I came across this website where different colormaps have been
compared
>> > >> and the author has come up with an optimal colormap for data
>> > >> visualization called the "cool-warm colormap".
>> > >>
>> > >> http://www.cs.unm.edu/~kmorel/documents/ColorMaps/index.html <
http://www.cs.unm.edu/%7Ekmorel/documents/ColorMaps/index.html>
>> > >>
>> > >> It is somewhat similar to the cool colormap already included in
>> > >> matplotlib, but I've added the new colormap to matplotlib in the
patch
>> > >> attached in case it is deemed fit to be included in the matplotlib
source.
>> > > We should include this, but I think the 257-entry version is
overkill;
>> > > it adds a big chunk to the _cm.py file, and I doubt it is visually
>> > > distinguishable from the 33-entry version. Would you mind providing
a
>> > > patch for the latter? (Or better yet, the functions that generate
the
>> > > r,g,b values.)
>> > Here's a pull request for the 33 entry map:
>> > <https://github.com/matplotlib/matplotlib/pull/486>
>>
>> Let me open a can of worms here...
>>
>> I looked at the paper, and the goal was specifically to produce a good
"default" colormap - not necessarily the best for any situation, but good
overall and certainly better than the rainbow ('jet') colormap in most
cases. (I agree with the author that jet is pretty terrible and tends to
distort data.)
>>
>> Should we switch to this as the default matplotlib colormap? I think it
would be a clear improvement.
>
> I have absolutely no clout here, but I'd definitely be in favor of
changing the default colormap away from "jet".
>
> Personally, I'd prefer a two-tone colormap as the default (two-distinct
tones at the limits with a gradient in-between---dubbed "sequential" in the
paper) instead of a three-tone colormap (three-distinct tones---dubbed
"diverging" in the paper). (I think this is a more common use case, and I
think using a "diverging" colormap effectively requires setting vmin/vmax.)
But really, (almost) anything is better than "jet".
>
> Don't misunderstand: I know I can change the default colormap in my
matplotlibrc file (and this is what I do). But 90% of people don't bother to
change the defaults. If changing this default in matplotlib prevents just 1
person from publishing a paper with a "jet" colormap, I think we'll have
made the world a better place. ;)
If it only prevents one person from publishing in "jet", then we really
stink in our jobs of promoting matplotlib...
Anyway, this is certainly is worthy of debate, but it certainly won't happen
for this release. We should be cutting RC tomorrow.
After the release, I encourage you guys to make your cases. Show us plots
that have been in "jet" and show them as better in another colormap.
As a bit of a challenge to you all, I am not color-blind, but I do wear
tinted glasses that make it difficult to tell the difference between darker
blues and black, and sometimes greens and blues are hard to distinguish.
 Furthermore, as a radar meteorologist, I am very accustomed to the
colormaps commonly used for radar reflectivies (and is similar to "jet").
Of course, I am not the only one to convince (and others could certainly
overrule me)...
Ben Root

On Thu, Sep 22, 2011 at 5:16 PM, Nathaniel Smith <nj...@po...> wrote:
> On Sep 21, 2011 5:29 PM, "Christoph Gohlke" <cg...@uc...> wrote:
> > On 9/13/2011 12:24 AM, Eric Firing wrote:
> > > On 07/18/2011 07:07 AM, Sameer Grover wrote:
> > >> I came across this website where different colormaps have been
> compared
> > >> and the author has come up with an optimal colormap for data
> > >> visualization called the "cool-warm colormap".
> > >>
> > >> http://www.cs.unm.edu/~kmorel/documents/ColorMaps/index.html
> > >>
> > >> It is somewhat similar to the cool colormap already included in
> > >> matplotlib, but I've added the new colormap to matplotlib in the patch
> > >> attached in case it is deemed fit to be included in the matplotlib
> source.
> > > We should include this, but I think the 257-entry version is overkill;
> > > it adds a big chunk to the _cm.py file, and I doubt it is visually
> > > distinguishable from the 33-entry version. Would you mind providing a
> > > patch for the latter? (Or better yet, the functions that generate the
> > > r,g,b values.)
> > Here's a pull request for the 33 entry map:
> > <https://github.com/matplotlib/matplotlib/pull/486>
>
> Let me open a can of worms here...
>
> I looked at the paper, and the goal was specifically to produce a good
> "default" colormap - not necessarily the best for any situation, but good
> overall and certainly better than the rainbow ('jet') colormap in most
> cases. (I agree with the author that jet is pretty terrible and tends to
> distort data.)
>
> Should we switch to this as the default matplotlib colormap? I think it
> would be a clear improvement.
>
I have absolutely no clout here, but I'd definitely be in favor of changing
the default colormap away from "jet".
Personally, I'd prefer a two-tone colormap as the default (two-distinct
tones at the limits with a gradient in-between---dubbed "sequential" in the
paper) instead of a three-tone colormap (three-distinct tones---dubbed
"diverging" in the paper). (I think this is a more common use case, and I
think using a "diverging" colormap effectively requires setting vmin/vmax.)
But really, (almost) anything is better than "jet".
Don't misunderstand: I know I can change the default colormap in my
matplotlibrc file (and this is what I do). But 90% of people don't bother to
change the defaults. If changing this default in matplotlib prevents just 1
person from publishing a paper with a "jet" colormap, I think we'll have
made the world a better place. ;)
-Tony

On Sep 21, 2011 5:29 PM, "Christoph Gohlke" <cg...@uc...> wrote:
> On 9/13/2011 12:24 AM, Eric Firing wrote:
> > On 07/18/2011 07:07 AM, Sameer Grover wrote:
> >> I came across this website where different colormaps have been compared
> >> and the author has come up with an optimal colormap for data
> >> visualization called the "cool-warm colormap".
> >>
> >> http://www.cs.unm.edu/~kmorel/documents/ColorMaps/index.html
> >>
> >> It is somewhat similar to the cool colormap already included in
> >> matplotlib, but I've added the new colormap to matplotlib in the patch
> >> attached in case it is deemed fit to be included in the matplotlib
source.
> > We should include this, but I think the 257-entry version is overkill;
> > it adds a big chunk to the _cm.py file, and I doubt it is visually
> > distinguishable from the 33-entry version. Would you mind providing a
> > patch for the latter? (Or better yet, the functions that generate the
> > r,g,b values.)
> Here's a pull request for the 33 entry map:
> <https://github.com/matplotlib/matplotlib/pull/486>
Let me open a can of worms here...
I looked at the paper, and the goal was specifically to produce a good
"default" colormap - not necessarily the best for any situation, but good
overall and certainly better than the rainbow ('jet') colormap in most
cases. (I agree with the author that jet is pretty terrible and tends to
distort data.)
Should we switch to this as the default matplotlib colormap? I think it
would be a clear improvement.
- Nathaniel

On 9/13/2011 12:24 AM, Eric Firing wrote:
> On 07/18/2011 07:07 AM, Sameer Grover wrote:
>> I came across this website where different colormaps have been compared
>> and the author has come up with an optimal colormap for data
>> visualization called the "cool-warm colormap".
>>
>> http://www.cs.unm.edu/~kmorel/documents/ColorMaps/index.html
>>
>> It is somewhat similar to the cool colormap already included in
>> matplotlib, but I've added the new colormap to matplotlib in the patch
>> attached in case it is deemed fit to be included in the matplotlib source.
>>
>> Regards,
>> Sameer
>
> Sameer,
>
> We should include this, but I think the 257-entry version is overkill;
> it adds a big chunk to the _cm.py file, and I doubt it is visually
> distinguishable from the 33-entry version. Would you mind providing a
> patch for the latter? (Or better yet, the functions that generate the
> r,g,b values.)
>
> Thank you.
>
> Eric
Here's a pull request for the 33 entry map:
<https://github.com/matplotlib/matplotlib/pull/486>
Christoph

On 9/19/2011 2:23 AM, Christoph Gohlke wrote:
>
>
> On 9/18/2011 2:30 PM, Eric Firing wrote:
>> On 09/18/2011 09:30 AM, Christoph Gohlke wrote:
>>> Hello,
>>>
>>> matplotlib uses int(x*255) or np.array(x*255, np.uint8) to quantize
>>> normalized floating point numbers x in the range [0.0 to 1.0] to
>>> integers in the range [0 to 255]. This way only 1.0 is mapped to 255,
>>> not for example 0.999. Is this really intended or would not the largest
>>> floating point number below 256.0 be a better scale factor than 255? The
>>> exact factor depends on the floating point precision (~255.999992 for
>>> np.float32, ~255.93 for np.float16).
>>>
>>> Christoph
>>
>> Christoph,
>>
>> It's a reasonable question; but do you have use cases in mind where it
>> actually makes a difference?
>>
>> The simple scaling with truncation is used in many places, both in the
>> python and the c++ code.
>>
>> Eric
>>
>
> Hi Eric,
>
> visually it will be hardly noticeable in most cases. However, I'd expect
> the histogram of normalized intensity data to be the same as the
> histogram of a linear grayscale image of that data (neglecting gamma
> correction, image scaling/interpolation for now). Consider this code for
> example:
>
> import numpy as np
> a = np.random.rand(1024*1024)
> a[0], a[-1] = 0.0, 1.0
> h0 = np.histogram(a, bins=256, range=(0, 1))[0]
> h1 = np.bincount(np.uint8(a * 255))
> h2 = np.bincount(np.uint8(a * 255.9999999999999))
> print (h0 - h1)
> print (h0 - h2)
>
> Christoph
>
To make this work with any float type one could use:
np.uint8(a * np.nextafter(a.dtype.type(256), a.dtype.type(0)))
Christoph

Paris (U.E.), le 21/09/2011
	Bonsoir
		I have been working in my infinitesimally small spare time for the last few months on triangle plots. Attached bellow are a few output examples (bulky because I chose the Tiff format for "historical" reasons). I still have quite a few variations I want to code (and perhaps an article to write dealing with triangle plots and some densities). I have not made anything public yet (bad habit, I know...), but it would be a shame if we duplicated work. Would you be interested in merging our work?
	Hubert Holin
On 20 sept. 2011, at 23:01, Kevin Davies wrote:
> Hello,
> I'm attaching two demo plots of a class that creates ternary or triangle plots. I've done the best I can, but I don't think the code is ready for prime time. I think it needs help from an expert on matplotlib projections. I've created a branch with the ternary.py file:
> https://github.com/kdavies4/matplotlib/compare/master...ternary
> 
> Kevin
> <ternary1.png><ternary2.png>------------------------------------------------------------------------------
> All the data continuously generated in your IT infrastructure contains a
> definitive record of customers, application performance, security
> threats, fraudulent activity and more. Splunk takes this data and makes
> sense of it. Business sense. IT sense. Common sense.
> http://p.sf.net/sfu/splunk-d2dcopy1_______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel

Sandro,
On Mon, Sep 19, 2011 at 4:08 PM, Sandro Tosi <mo...@de...> wrote:
> Moreover, what are the
> sources of those files (i.e. where are they downloaded from)?
>
>
>From the README,
the coastline, lake, river and political boundary data are extracted
from datasets provided with the Generic Mapping Tools
(http://gmt.soest.hawaii.edu)
and are included under the terms given in LICENSE_data.
a public domain 5-minute land/sea/lake mask dataset from
http://www.ngdc.noaa.gov/seg/cdroms/graham/graham/graham.htm
is included.
Ben Root

Hello,
I would like to ask for a code review on the Sankey feature. I've
 tried to improve the formatting and documentation. Here is the
 comparison:
https://github.com/kdavies4/matplotlib/compare/master...sankey3#diff-1
Thanks!
Kevin

Hello,
To be available from the Debian main archive, we need to provide a
tarball containing all source code, something that later generates a
"binary" content or in a form that's what called "the preferred form
of modification" (f.e. in a sphinx doc, it's the rst files, not the
html files resulting from the building process, since the preferred
form of modification of that documentation is thru the rst files).
The question comes directly to the datafile: is the file format
datafiles are shipped the preferred form of modification? if yes, what
are the ways/tools to modify those files? Moreover, what are the
sources of those files (i.e. where are they downloaded from)?
I know that it might seem a bit picky, from a Debian POV, these
information are really important.
Thanks & Cheers,
-- 
Sandro Tosi (aka morph, morpheus, matrixhasu)
My website: http://matrixhasu.altervista.org/
Me at Debian: http://wiki.debian.org/SandroTosi

On 9/19/11 2:54 PM, Sandro Tosi wrote:
> Hello,
> it seems nad2bin is not installed, but only compiled. Is that
> expected? Should it be installed by hand instead of by setup.py
> install?
>
> Cheers,
It does not need to be installed.
-Jeff
-- 
Jeffrey S. Whitaker Phone : (303)497-6313
Meteorologist FAX : (303)497-6449
NOAA/OAR/PSD R/PSD1 Email : Jef...@no...
325 Broadway Office : Skaggs Research Cntr 1D-113
Boulder, CO, USA 80303-3328 Web : http://tinyurl.com/5telg

Hello,
it seems nad2bin is not installed, but only compiled. Is that
expected? Should it be installed by hand instead of by setup.py
install?
Cheers,
-- 
Sandro Tosi (aka morph, morpheus, matrixhasu)
My website: http://matrixhasu.altervista.org/
Me at Debian: http://wiki.debian.org/SandroTosi

Hello,
the doc build process is a lot of hand-crafting work and it also
requires the full source of matplotlib just to get a couple of
directory.
I know that code duplication sucks, but I'd suggest to include
directly into the basemap dir all the files (sphinxext & _static at
the very least) needed for the doc compilation. This would help a lot
the distributions (Debian of course :) ).
Thanks for considering,
-- 
Sandro Tosi (aka morph, morpheus, matrixhasu)
My website: http://matrixhasu.altervista.org/
Me at Debian: http://wiki.debian.org/SandroTosi

On 9/18/11 8:49 AM, Sandro Tosi wrote:
> Hi,
> when running
>
> python setup.py clean
>
> nad2bin is compiled. I've just worked around with the attached patch,
> so it would be nice if you can integrate it upstream or come up with a
> better solution.
>
> Regards,
Applied your patch to github master. Will be in 1.0.2, which I will 
release soon.
-Jeff
-- 
Jeffrey S. Whitaker Phone : (303)497-6313
Meteorologist FAX : (303)497-6449
NOAA/OAR/PSD R/PSD1 Email : Jef...@no...
325 Broadway Office : Skaggs Research Cntr 1D-113
Boulder, CO, USA 80303-3328 Web : http://tinyurl.com/5telg

On 9/18/2011 2:30 PM, Eric Firing wrote:
> On 09/18/2011 09:30 AM, Christoph Gohlke wrote:
>> Hello,
>>
>> matplotlib uses int(x*255) or np.array(x*255, np.uint8) to quantize
>> normalized floating point numbers x in the range [0.0 to 1.0] to
>> integers in the range [0 to 255]. This way only 1.0 is mapped to 255,
>> not for example 0.999. Is this really intended or would not the largest
>> floating point number below 256.0 be a better scale factor than 255? The
>> exact factor depends on the floating point precision (~255.999992 for
>> np.float32, ~255.93 for np.float16).
>>
>> Christoph
>
> Christoph,
>
> It's a reasonable question; but do you have use cases in mind where it
> actually makes a difference?
>
> The simple scaling with truncation is used in many places, both in the
> python and the c++ code.
>
> Eric
>
Hi Eric,
visually it will be hardly noticeable in most cases. However, I'd expect 
the histogram of normalized intensity data to be the same as the 
histogram of a linear grayscale image of that data (neglecting gamma 
correction, image scaling/interpolation for now). Consider this code for 
example:
import numpy as np
a = np.random.rand(1024*1024)
a[0], a[-1] = 0.0, 1.0
h0 = np.histogram(a, bins=256, range=(0, 1))[0]
h1 = np.bincount(np.uint8(a * 255))
h2 = np.bincount(np.uint8(a * 255.9999999999999))
print (h0 - h1)
print (h0 - h2)
Christoph