matplotlib-users

[Matplotlib-users] pcolor problem

[burak n. <bur...@gm...>]

Hello everone,

I am trying to use pcolor from pylab but I can't get it work. As you know
signature of pcolor for Matlab is pcolor(x,y,data) which x and y are
vectors. When I try pcolor(data)

it works but it is not what I want. Because without x,y vectors placed it
is not generating correct result. when i put x and y vectors in place I get
a ValueError: shape mismatch: objects cannot be broadcast to a single shape

but in matlab it is working with x , y vectors.

Any idea.

Thanks a lot for your time


pcolor(X,Y,reshaped)  ----> reshaped is my data matrix
show()


Burak

Re: [Matplotlib-users] pcolor problem

[Eric F. <ef...@ha...>]

On 2013/01/19 3:59 AM, burak nebioğlu wrote:
> Hello everone,
>
> I am trying to use pcolor from pylab but I can't get it work. As you
> know signature of pcolor for Matlab is pcolor(x,y,data) which x and y
> are vectors. When I try pcolor(data)
>
> it works but it is not what I want. Because without x,y vectors placed
> it is not generating correct result. when i put x and y vectors in place
> I get a ValueError: shape mismatch: objects cannot be broadcast to a
> single shape

Right, because your x, y, and data don't have properly matching shapes.

http://matplotlib.org/api/axes_api.html?highlight=pcolor#matplotlib.axes.Axes.pcolor

The various ways to make them match are explained in the link above.

Note also that in most cases, using pcolormesh provides the desired 
result but with a much faster implementation.  Pcolor is slow, and 
should be used only for rather small arrays.

Eric


>
> but in matlab it is working with x , y vectors.
>
> Any idea.
>
> Thanks a lot for your time
>
>
> pcolor(X,Y,reshaped)  ----> reshaped is my data matrix
> show()
>
>
> Burak
>
>
> ------------------------------------------------------------------------------
> Master Visual Studio, SharePoint, SQL, ASP.NET, C# 2012, HTML5, CSS,
> MVC, Windows 8 Apps, JavaScript and much more. Keep your skills current
> with LearnDevNow - 3,200 step-by-step video tutorials by Microsoft
> MVPs and experts. SALE $99.99 this month only -- learn more at:
> http://p.sf.net/sfu/learnmore_122912
>
>
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

Re: [Matplotlib-users] pcolor problem

[Eric F. <ef...@ha...>]

On 2013/01/20 1:48 AM, Burak nebioğlu wrote:
> Dear Eric ,
>
> below are the results that I get when I see these errors it is also
> strange for me but I tried in matlab and it worked with mtlab style I
> tried for pcolormesh it works just like matlab it is (99,) (56,)
>   dimensioned 2 vectors and data is 99*56 matrix

Sorry, I included the () by mistake on the requested dtype and the shape 
print statements.  dtype and shape are attributes, so they don't need 
parentheses. In any case, you provided enough information for me to 
test, and I see exactly what you note--except that the bug is in 
pcolormesh, not pcolor.  With either one, an argument sequence (x, y, z) 
should require that the shape of z be (ny, nx) or (ny-1, nx-2), not the 
transpose of that.  If your present data array has shape (nx, ny), then use

pcolormesh(x, y, data.T)

and it will be plotted correctly.

It may be that matlab is trying to be helpful by doing a transpose 
automatically if that makes the function arguments seem to fit better, 
but that is not what we do, or what we would ever want to do.  So, 
regardless of whether you use pcolor or pcolormesh, please use the 
argument order and shape requirements as described in the docstrings, 
the API documentation, and the examples.  Otherwise you are not likely 
to get the result you want.

I have filed a bug report for this:
https://github.com/matplotlib/matplotlib/issues/1688

Eric

>
>
>
>
> type x  <type 'numpy.ndarray'>
> X.dtype
>
>      print 'X.dtype',X.dtype()
> TypeError: 'numpy.dtype' object is not callable
>
>      print 'X.shape',X.shape()
> TypeError: 'tuple' object is not callable
>
>
> type y  <type 'numpy.ndarray'>
> y.shape
>
>      print 'y.dtype',Y.dtype()
> TypeError: 'numpy.dtype' object is not callable
>
>
>      print 'y.shape',Y.shape()
> TypeError: 'tuple' object is not callable
>
> type data  <type 'numpy.ndarray'>
> data.dtype
>
>      print 'data.dtype',reshaped.dtype()
> TypeError: 'numpy.dtype' object is not callable
>
>      print 'data.shape',reshaped.shape()
> TypeError: 'tuple' object is not callable
>
> Thank you again
>
>
>
>
> 2013/1/20 Eric Firing <ef...@ha... <mailto:ef...@ha...>>
>
>     print type(x)
>     print x.dtype()
>     print x.shape()
>
>
>