matplotlib-users Mailing List for matplotlib

Thank you very much, you helped me a great deal. The video tutorials are great.

I'm going through the stuff you sent and things seem to become clearer.... 

Tomislav

----- Original Message -----
From: John Hunter
Sent: 04/11/10 04:19 PM
To: tom...@gm...
Subject: Re: [Matplotlib-users] how to plot a Polygon / plt.draw() problem

On Sun, Apr 11, 2010 at 7:15 AM, tom...@gm...
<tom...@gm...> wrote:
> can someone help me to plot a polygon in matplotlib?
>
> I have been reading about the axes.patches.Polygon class and I have defined
> the
>
> Polygon object that has a preset lw and points. How do I plot it?
>
> I'm confused because the Axes documentation states that this class holds
> most of
>
> the figure objects like Rectangle, Line2D, and then the website states that
> the Line2D
>
> is a return object from the plt.plot() invocation.

Yes, Axes.plot is a helper function which creates a Line2D object,
adds it to the axes, sets the transformation, etc... This process is
covered in some detail in the matplotlib Artist tutorial

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

and in the advanced matplotlib tutorial at scipy -- video available here

 http://www.archive.org/details/scipy09_advancedTutorialDay1_3

> What if I create my own
> set of Rectangle
>
> (Polygon) objects and want to create a list of them and plot them?

If you create your own polygons/rectangles/patches, create them, and
then add them with Axes.add_patch

 http://matplotlib.sourceforge.net/api/axes_api.html#matplotlib.axes.Axes.add_patch

If you want to create a bunch of them, consider a PolygonCollection
(or a RegularPolygonCollection depending on your use case)

 http://matplotlib.sourceforge.net/api/collections_api.html
 http://matplotlib.sourceforge.net/search.html?q=codex+PolyCollection

> Also, I'm using this sequence of commands to work in OO mode interactively
>
> (just to learn) but when I execute plt.draw() no figure appears.

We make a distinction between raising a figure (plt.show) and
rendering to an existing figure (plt.draw). In interactive mode
(which is what ipython -pylab turns on) figures are automatically
raised/shown. You can control these settings from a regular python
shell using ion and ioff. See

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

Here is a complete example::

 import matplotlib.patches as patches
 import matplotlib.pyplot as plt

 fig = plt.figure()
 ax = fig.add_subplot(111)

 verts = [0,0], [0,1], [1,1], [1,0]

 poly = patches.Polygon(verts)

 ax.add_patch(poly)

 ax.set_xlim(-2,2)
 ax.set_ylim(-2,2)

 plt.show()


Hope this helps,
JDH

I need to plot the path the tip of a specific vector draws,
when the vector is rotated around certain axis for some angels

as a basic example:
start vector is (0,0,1)
rotation around axis (0,1,0) for an angle of (pi/2)
rotation around axis (0,0,1) for an angle of (pi)

so the path should show something like starting at the bottom,
go up to the equator plane and then turn around the equator
to the other side

this could also be expressed in terms of a point moving on
a sphere ...

a graphical example would be like the path C in this figure
http://physics.aps.org/view_image/2213/medium/4

does anyone happen to know how i could easily implement something
like that?

with best regards
and thanks for your help/suggestions
  q

-- 
The king who needs to remind his people of his rank, is no king.

A beggar's mistake harms no one but the beggar. A king's mistake,
however, harms everyone but the king. Too often, the measure of
power lies not in the number who obey your will, but in the number
who suffer your stupidity.

Firstly, I want to thank Marius (Marius 't Hart-3) and Eric (efiring) for the
help on my question. Especially Eric, your advices about 'antialiased' gave
me the hint to get the solution.

In my previous topic "
http://old.nabble.com/contourf-creats-white-like-lines-(or-gaps)-between-each-two-color-patches-tp27982822p28210898.html
contourf creats white-like lines (or gaps) between each two color patches ",
I attached two figures to show the bug-like problem of contourf plotting
function. The first one is created by contourf only, and the in second one,
I used contourf and contour together to fill up the gaps between each two
patches.
It seems a kind of solution to make the filled contour map better looking,
but I really don't like this messy way. So I posted the questionhere to find
answers.

Following Eric's suggestion
> CS = contourf(Z)
> for c in CS.collections:
>     c.set_antialiased(False)
the antialiased attributes of each polygon in the filled contour map is
changed to be unsmoothed, and the gaps are gone. This is much better way to
solve the gap problem, but I should do this every time after I plotting the
filled contour map. It's not the best solution.

1. Basic solution

I tried to study the arguments of contourf function to find the proper
parameter about antiliased, and there is one in the doc string:
'''
    contourf-only keyword arguments:
    
      *antialiased*: [ True | False ]
        enable antialiasing
'''
Good! Try this:
                CS = contourf(Z, antialiased=False)
Oops~~~ nothing happens, gaps are still there. Why? Bug?

I searched 'antialiased' in the contour.py(this file located in the
installed matplotlib package directory), and find this: (line 604 through
607)
                col = collections.PolyCollection(nlist,
                                     antialiaseds = (self.antialiased,),
                                     edgecolors= 'none',
                                     alpha=self.alpha)
I changed the line 605
                                     antialiaseds = (self.antialiased,),
to
                                     antialiaseds = self.antialiased,
plotted the filled contourf map again:
                CS = contourf(Z, antialiased=False)
Wow~~~ gaps free.

I don't know if it is a bug or not, but edit it simply, works well.


2. Upgrade the solution

Don't like the unsmoothed edge? Still want the antialiased to be True?
OK, notice the line just under antialiaseds setting listed above?
                                    edgecolors='none',
it means no edge is plotted.

Go straight down to line 640:
                collection.set_facecolor(color)
here the face colors are set. Just below this line, I added one new line:
                collection.set_edgecolor(color)
then plotted the filled contourf map anain, without setting
antialiased=False:
                CS = contourf(Z)
Aha, Mission accomplished.

-- 
View this message in context: http://old.nabble.com/Solution-of-my-previous-topic%3A-contourf-creats-white-like-lines-%28or-gaps%29-tp28219185p28219185.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

2010/4/11 Alan G Isaac <ala...@gm...>:
> On 4/11/2010 9:27 AM, Friedrich Romstedt wrote:
>> I think you can use Tk via the Tkinter Python package.  On linux I
>> heard it's looking a bit weird, but as a starting points it's easy
>> enough.
>
> Weird how?
> Will that be fixed with the new release (ttk, in Python 2.7)?

Hmm, it was meant as a warning, nothing shure.  I just read somewhere
that the Tk, although originating from linux, would look the worst on
linux.  I can neither verify nor falsify, because I have no linux.

Sorry for the noise.  I even cannot exclude that it's wrong.

This amazing page: http://docs.python.org/dev/library/ttk.html (I
especially like the ComboBox, something I always missed in Tkinter :-)
states that look and feed would change with ttk, so maybe you're
lucky.

Friedrich

My simple way to solve the white-lines issue is simply to give the contourf
command twice:

contourf(x,y,z)
contourf(x,y,z)

The white lines are magically disappearing, also in png files.
This works well also with alpha<1

Tsviki Hirsh


> What version of mpl are you using?
My mpl version is 0.99.1

> Are you modifying the default anti-aliasing in the patch collections that
> contour is creating?
Could you please tell me how to this? I am very new to matplotlib, thank you
very much if you can give me some advices.

>Are you seeing the problem when using the ps and pdf backends, or only with
agg (that is, creating png files directly)?
So far I only creating png files directly. Also, how to use ps and pdf
backend?

> ....which has been fixed in svn.
Do you mean I should download another mpl by svn?

Thank you for your answers very much!


efiring wrote:
>
> What version of mpl are you using?  Are you modifying the default
> anti-aliasing in the patch collections that contour is creating? Are you
> seeing the problem when using the ps and pdf backends, or only with agg
> (that is, creating png files directly)?
>
> There are two problems that can contribute to this, one related to the
> way the agg backend handles the boundaries between filled regions, and
> the other a bug in path simplification, which has been fixed in svn.
>
> Eric
>
>
>
------------------------------------------------------------------------------
> Download Intel&#174; Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>
>

--
View this message in context:
http://old.nabble.com/contourf-creats-white-like-lines-%28or-gaps%29-between-each-two-color-patches-tp27982822p28210898.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

I figured it out on my own! Why I couldn't figure it out before... I don't
know...
Here is the code:

myhighlights[1].set_facecolor('y')
myhighlights[1].set_xy([[ 3.5 ,  0.  ],\
    [ 3.5 ,  1.  ],
    [ 3.75,  1.  ],
    [ 3.75,  0.  ],
    [ 3.5 ,  0.  ]])


KrishnaPribadi wrote:
> 
> Hi,
> I added several axvspan "highlights on my plot.
> 
> I'd like to change the properties of only one of those "highlights" such
> as the facecolor, xmin, and xmax. I've included some code below. Lets say
> I'd like to change the properties on the 2nd vertical highlight (axvspan).
> 
> Can someone please suggest some code on accessing and changing properties
> on that specific object?
> 
> do I access it from the object "myhighlights" or do I access is from the
> object "ax"?
> 
> Note: This is just some simple code I came up with. My real application is
> in a wx app GUI so I will be using the draw() function.
> 
> Here is my code:
> import numpy as np
> import matplotlib.pyplot as plt
> 
> 
> t = np.arange(0, np.pi*2, 0.01)
> x = np.sin(2*np.pi*t)
> 
> 
> fig = plt.figure()
> ax = fig.add_subplot(111)
> ax.grid('On')
> 
> myplot = ax.plot(t, x, 'b')
> 
> myhighlights = []
> myhighlights.append(ax.axvspan(1,2, facecolor='g', alpha=0.2))
> myhighlights.append(ax.axvspan(3,4, facecolor='g', alpha=0.2)) #later
> let's change the facecolor, xmin, and xmax
> myhighlights.append(ax.axvspan(5,6, facecolor='g', alpha=0.2))
> 
> #need code to change properties of the 2nd axvspan
> #change facecolor to yellow
> #change xmin to 3.5 and xmax to 4.5
> 
> plt.show()
> 


-----
Krishna Adrianto Pribadi
Test Engineer
Harley-Davidson Motor Co.
Talladega Test Facility
Vehicle Test Stands
-- 
View this message in context: http://old.nabble.com/axvspan%2C-multiple-calls%2C-how-to-change-prop-of-1-object-instance-tp28219155p28219166.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

Hi,
I added several axvspan "highlights on my plot.

I'd like to change the properties of only one of those "highlights" such as
the facecolor, xmin, and xmax. I've included some code below. Lets say I'd
like to change the properties on the 2nd vertical highlight (axvspan).

Can someone please suggest some code on accessing and changing properties on
that specific object?

do I access it from the object "myhighlights" or do I access is from the
object "ax"?

Note: This is just some simple code I came up with. My real application is
in a wx app GUI so I will be using the draw() function.

Here is my code:
import numpy as np
import matplotlib.pyplot as plt


t = np.arange(0, np.pi*2, 0.01)
x = np.sin(2*np.pi*t)


fig = plt.figure()
ax = fig.add_subplot(111)
ax.grid('On')

myplot = ax.plot(t, x, 'b')

myhighlights = []
myhighlights.append(ax.axvspan(1,2, facecolor='g', alpha=0.2))
myhighlights.append(ax.axvspan(3,4, facecolor='g', alpha=0.2)) #later let's
change the facecolor, xmin, and xmax
myhighlights.append(ax.axvspan(5,6, facecolor='g', alpha=0.2))

#need code to change properties of the 2nd axvspan
#change facecolor to yellow
#change xmin to 3.5 and xmax to 4.5

plt.show()

-----
Krishna Adrianto Pribadi
Test Engineer
Harley-Davidson Motor Co.
Talladega Test Facility
Vehicle Test Stands
-- 
View this message in context: http://old.nabble.com/axvspan%2C-multiple-calls%2C-how-to-change-prop-of-1-object-instance-tp28219155p28219155.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

> CS = contourf(Z)
> for c in CS.collections:
>     c.set_antialiased(False)

Eric,
Thank you! I like this very much!
So, the last thing is, is there any way to set this antialiased feature to
False by default?


efiring wrote:
> 
> lmkli wrote:
>>> What version of mpl are you using?  
>> My mpl version is 0.99.1
>> 
>>> Are you modifying the default anti-aliasing in the patch collections
>>> that
>>> contour is creating?
>> Could you please tell me how to this? I am very new to matplotlib, thank
>> you
>> very much if you can give me some advices.
> 
> see below.
> 
>> 
>>> Are you seeing the problem when using the ps and pdf backends, or only
>>> with
>> agg (that is, creating png files directly)?
>> So far I only creating png files directly. Also, how to use ps and pdf
>> backend?
> 
> The easiest way is with the pyplot savefig command: e.g.
> 
> savefig("test.pdf")
> 
> to save the current figure as a pdf.
> 
> It doesn't look to me like this solves the problem, though.
> 
>> 
>>> ....which has been fixed in svn.
>> Do you mean I should download another mpl by svn?
> 
> That might help, but it probably won't cure the problem completely.  I 
> don't know that there is any combination of parameters that gives 
> boundaries as smooth and gap-free as in your second example.  You may 
> get a bit closer to it like this:
> 
> CS = contourf(Z)
> for c in CS.collections:
>      c.set_antialiased(False)
> 
> (If working interactively in ipython, follow this with a "draw()" 
> command to force a redraw.)
> 
> Eric
> 
>> 
>> Thank you for your answers very much!
>> 
>> 
>> efiring wrote:
>>> What version of mpl are you using?  Are you modifying the default 
>>> anti-aliasing in the patch collections that contour is creating? Are you 
>>> seeing the problem when using the ps and pdf backends, or only with agg 
>>> (that is, creating png files directly)?
>>>
>>> There are two problems that can contribute to this, one related to the 
>>> way the agg backend handles the boundaries between filled regions, and 
>>> the other a bug in path simplification, which has been fixed in svn.
>>>
>>> Eric
>>>
>>>
>>> ------------------------------------------------------------------------------
>>> Download Intel® Parallel Studio Eval
>>> Try the new software tools for yourself. Speed compiling, find bugs
>>> proactively, and fine-tune applications for parallel performance.
>>> See why Intel Parallel Studio got high marks during beta.
>>> http://p.sf.net/sfu/intel-sw-dev
>>> _______________________________________________
>>> Matplotlib-users mailing list
>>> Mat...@li...
>>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>>
>>>
>> 
> 
> 
> ------------------------------------------------------------------------------
> Download Intel® Parallel Studio Eval
> Try the new software tools for yourself. Speed compiling, find bugs
> proactively, and fine-tune applications for parallel performance.
> See why Intel Parallel Studio got high marks during beta.
> http://p.sf.net/sfu/intel-sw-dev
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
> 

-- 
View this message in context: http://old.nabble.com/contourf-creats-white-like-lines-%28or-gaps%29-between-each-two-color-patches-tp27982822p28219132.html
Sent from the matplotlib - users mailing list archive at Nabble.com.

Thanks for point TKinter to me. However, I'm stuck again.

I've tried two approaches, one is following what you suggested:

""" Tkinter  """
import Tkinter as tk
root = tk.Tk()
from PIL import Image, ImageTk
image = ImageTk.PhotoImage(Image.open('map.png')) # load saved image
#image = ImageTk.PhotoImage(im) # load image from StringIO
tk.Label(root, image=image).pack()

The other is to convert the PIL image to array and show with imshow()

""" PIL to array """
from matplotlib.image import pil_to_array
rgba = pil_to_array(Image.open('map.png')) # load saved image
#rgba = pil_to_array(im) # load image from StringIO
rgba = rgba.astype(np.float32)/255.
imshow(rgba)

Both work fine for a saved image. However, they return strange errors
when I tried to apply them to the image from "StringIO", or even for a
saved image in the same script used to generate them. I guess that
something I import from matplotlib is causing a conflict here.

Here are the error messages:

""" Tkinter  """
_tkinter.TclError: image "pyimage9" doesn't exist


""" PIL to array """
terminate called after throwing an instance of 'char const*'

I'm attaching the script.
Aborted

There doesn't seem to be a screenshot attached.  Maybe the mailing list 
ate it.  Can you send it to me directly (off list)?

Mike

KrishnaPribadi wrote:
> Here my screen shot. Notice the red and green line merging then turning
> brown...
>
>   

-- 
Michael Droettboom
Science Software Branch
Operations and Engineering Division
Space Telescope Science Institute
Operated by AURA for NASA

lmkli wrote:
>> What version of mpl are you using?  
> My mpl version is 0.99.1
> 
>> Are you modifying the default anti-aliasing in the patch collections that
>> contour is creating?
> Could you please tell me how to this? I am very new to matplotlib, thank you
> very much if you can give me some advices.

see below.

> 
>> Are you seeing the problem when using the ps and pdf backends, or only with
> agg (that is, creating png files directly)?
> So far I only creating png files directly. Also, how to use ps and pdf
> backend?

The easiest way is with the pyplot savefig command: e.g.

savefig("test.pdf")

to save the current figure as a pdf.

It doesn't look to me like this solves the problem, though.

> 
>> ....which has been fixed in svn.
> Do you mean I should download another mpl by svn?

That might help, but it probably won't cure the problem completely.  I 
don't know that there is any combination of parameters that gives 
boundaries as smooth and gap-free as in your second example.  You may 
get a bit closer to it like this:

CS = contourf(Z)
for c in CS.collections:
     c.set_antialiased(False)

(If working interactively in ipython, follow this with a "draw()" 
command to force a redraw.)

Eric

> 
> Thank you for your answers very much!
> 
> 
> efiring wrote:
>> What version of mpl are you using?  Are you modifying the default 
>> anti-aliasing in the patch collections that contour is creating? Are you 
>> seeing the problem when using the ps and pdf backends, or only with agg 
>> (that is, creating png files directly)?
>>
>> There are two problems that can contribute to this, one related to the 
>> way the agg backend handles the boundaries between filled regions, and 
>> the other a bug in path simplification, which has been fixed in svn.
>>
>> Eric
>>
>>
>> ------------------------------------------------------------------------------
>> Download Intel® Parallel Studio Eval
>> Try the new software tools for yourself. Speed compiling, find bugs
>> proactively, and fine-tune applications for parallel performance.
>> See why Intel Parallel Studio got high marks during beta.
>> http://p.sf.net/sfu/intel-sw-dev
>> _______________________________________________
>> Matplotlib-users mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>>
>>
>