[chimera-dev] colors...

[Eric Pettersen]

On Friday, September 26, 2003, at 02:41 AM, Lars Kunert wrote:

> Hi I do like to color some molecules/atoms
>
> I figured out, that I create a color like this:
>
>>> c=chimera.Color()
>>> c=(1,0,0,1)

Does the code above have a typo in it?  The first line assigns 'c' to  
point at a Color instance, then the second line has 'c' point to a  
four-tuple (effectively discarding the Color instance).  Perhaps you  
meant the second line to be 'c.rgba = (1,0,0,1)'?

Regardless of the 'c' assignment problem, the more basic problem is  
that Color is a base class (for MaterialColor and TextureColor) that  
isn't supposed to be used directly like this.  For the most part you  
want to use MaterialColors when assigning colors in Chimera.  To get  
the color you wanted above, there are two methods:

	from chimera.colorTable import getColorByName
	c = getColorByName("red")

The getColorByName() function returns a MaterialColor instance.  You  
can use any of the color names in the Actions...Color menu with  
getColorByName (e.g. "dodger blue").

If you need a shade that isn't one of the built-in color names, you  
would have to do something similar to what you were trying to do  
earlier:

	c = chimera.MaterialColor(1, 0, 0, None)  # 'None' is the "material",  
i.e. use the default material

> But when I apply this to all the atoms of a molecule, I end up with  
> the following error-message:
>>>> for a in m.atoms:
>         a.color=c
>
> Traceback (most recent call last):
>   File "<pyshell#66>", line 1, in ?
>     for a in m.atoms:
> RuntimeError: __wrappy__ is missing
>>>>

I don't get this error using the code you supplied.  In both the case  
where 'c' is a four-tuple and the case where it's a Color instance, I  
get:

TypeError: argument 1 should be a const Color*

In your code, is 'm' a Molecule instance created by  
chimera.openModels.open(), or is it something else?

> I suppose, that colors like "red"are predefined somewhere, where?

Yes, in the chimera.colorTable module, as in the example code above.

> Are the different color-levels modell/molecule/atom are just fancy  
> stuff for the interactive user? - I colored a molecule first  
> (interactive) "byatom" and then tried to alter the coloring by hand:
>>> m.color=c
>
> it did not work, I had to alter the color atomwise... is color-byatom  
> a function? How can I call it?

Coloring "by atom" assigns colors to individual atoms, which overrides  
colors assigned at the model level, which is why your color assignment  
to the molecule had no apparent effect.  The is explained in detail in  
a new programmer's example (i.e. it's not in the 1700 release) called  
"Chimera's Object Model" which explains not only the color hierarchy  
(in the section "Display Properties") but the whole basic layout of  
Chimera objects.  To access this example you need to go to  
http://www.cgl.ucsf.edu/chimera/docs/ProgrammersGuide/Examples/ 
index.html (it's the first example).

There are two functions to assist in coloring by element type.  They  
are both in the chimera.actions module.  The first is colorByElement(),  
which colors atoms in the current selection, or everything if nothing  
is selected.  By default it sets the "color" attribute (the atom's  
color).  You can supply the keyword 'attr="surfaceColor"' to make it  
color surfaces instead, or 'attr="labelColor"' to make it color labels.  
  Perhaps more useful is the colorAtomByElement(a) function which takes  
an atom as argument.  It takes the same 'attr' keyword that  
colorByElement() does.  In the next release (in about two weeks), it  
will instead take an 'attrs' keyword for a list of attributes to set.

> Is there a way to "switch on"a history function in the buildin IDLE?  
> Copy and paste do also not work within the IDLE. I allways have to  
> copy and paste via an editor...

In IDLE, you can insert an old line of code by clicking the mouse to  
put the insertion point in that line of code and then hitting Return.   
The code will appear at the bottom and you can edit it and so forth.


                         Eric Pettersen
                         UCSF Computer Graphics Lab
                         pett at cgl.ucsf.edu
                         http://www.cgl.ucsf.edu