Re: [Matplotlib-users] Reliefplot (imshow with shading)

Timothée Lecomte wrote:
> Jeff Whitaker wrote:
>> Jeff Whitaker wrote:
>>> Timothée Lecomte wrote:
>>>> Dear all,
>>>>
>>>> I am using matplotlib with a great pleasure, and I enjoy its 
>>>> capabilities.
>>>> I have recently attended a conference where the invited speaker 
>>>> showed great visualizations of arrays from both experiments and 
>>>> simulations. His plots were basically looking like those produced 
>>>> by imshow, that is a luminance array rendered as a colormap image, 
>>>> but with the additionnal use of a shading, which gives a really 
>>>> great feeling to the image. You can feel the height of each part of 
>>>> the image.
>>>>
>>>> I have tried to find what software could have produced such a plot, 
>>>> and found the ReliefPlot function of Mathematica, which has 
>>>> precisely this purpose : rendering a colormap image from an array 
>>>> with a shading to give the perception of relief.
>>>>
>>>> The documentation and its examples are self-explanatory :
>>>> http://reference.wolfram.com/mathematica/ref/ReliefPlot.html
>>>> (look in particular at the first "neat example" at the bottom of 
>>>> that page)
>>>>
>>>> The two "live" demonstrations illustrate this plot style quite well 
>>>> too :
>>>> http://demonstrations.wolfram.com/ReliefShadedElevationMap/
>>>> http://demonstrations.wolfram.com/VoronoiImage/
>>>>
>>>> So here are my questions :
>>>> Is there a trick to generate an image with such a shading in 
>>>> matplotlib ?
>>>> If not, do you know of a python tool that could help ?
>>>> Where could I start if I want to code it myself in matplotlib ?
>>>>
>>>> Thanks for your help.
>>>>
>>>> Best regards,
>>>>
>>>> Timothée Lecomte
>>>>
>>>>   
>>>
>>> Timothée:  There is nothing built-in, but it would be a nice thing 
>>> to have.  Here's a proof-of-concept hack that follows the approach 
>>> used in the Generic Mapping Tools (explained here 
>>> http://www.seismo.ethz.ch/gmt/doc/html/tutorial/node70.html), with 
>>> some code borrowed from http://www.langarson.com.au/blog/?p=14.  
>>> It's very rough, but if it looks promising to you I can try to 
>>> polish it.
>>>
>>> -Jeff
>>
>> Found a bug, here's a fixed version.
>>
>> -Jeff
>>
> Hi Jeff,
>
> Sure it looks promising ! The example you provided is very nice. I 
> will try on my own data on Monday, and I'll let you know if it gives a 
> good result too. Thank you very much for that very fast hack !
>
> Best regards,
>
> Timothée
>
>
Timothée:  I've added this capability in svn, along with an example 
(shading_example.py) to show how to use it.  Thanks for suggesting it.

-Jeff