On Fri, Feb 6, 2009 at 6:38 PM, Sébastien Morin
<sebastien.morin.1@xxxxxxxxx> wrote:
Hi,
I was looking at the code for function 'vectors()' in both
'prompt/structure.py' and 'generic_fns/structure/main.py' and saw some
possible problems...
============
1.
============
The default arguments for 'prompt/structure.vectors()' are:
attached='H'
spin_id=None
model=None
verbosity=1
ave=True
unit=True
while, for 'generic_fns/structure/main.vectors()', they are:
attached=None
spin_id=None
model=None
verbosity=1
ave=True
unit=True
The attached argument has two different default values ? Is that normal ?
This is perfectly normal. The vectors() function in
generic_fns.structure.main is supposed to be generic for calculating
the vector between 2 atoms. The structure.vectors() user function is
used more often in relax scripts for getting the XH bond vector where
X is represented by the spin container and H is the attached atom.
'H' probably the most generic name for the attached proton, and this
works nicely for protein NH work. Note this default value is used
simply as a suggestion and if a user forgets to set this correctly,
they will quickly find out with all the warning messages.
============
2.
============
There seems to be an error in the docstring example for
'prompt/structure.vectors()' where the argument 'attached' is replaced
by 'proton'...
relax> structure.vectors(proton='H', spin_id='@N')
...
relax> structure.vectors(proton='HN', spin_id='@N')
...
relax> structure.vectors(proton='HA', spin_id='@CA')
Should this be corrected ?
Yes! The 'attached_atom' argument used to be called 'proton' until I
generalised the user function. So this is just a out of date
docstring.
Cheers,
Edward
Re: structure.vector()