Hi again, 

My mistake, 
the number of point grids in the ouput dx file (81 108 108) allow to recreate the complete grid... 

So If I understand well how the async (or parallel) is working...
 The calculation is performed with a grid of 97x129X129 and only with 81x108x108 points (error in my previous message) are written in the dx file.
(The extra points come form the ofrac keyword ?)

So in this case, the focusFillBound warning  is not harmful ?
Thanks a lot.

Nicolas Coquelle.

Début du message réexpédié :

De : Nicolas Coquelle <coquelle@ualberta.ca>
Date : 20 mai 2010 12:33:05 HAR
À : APBS Users Mailing List <apbs-users@lists.sourceforge.net>
Objet : [Apbs-users] Async calculation

Hi APBS Users,

To get a high resolution potentiel map, I was trying to use the async  
method (running APBS 1.2.1b)
I genereated input files using the inputgen.py scripts and ended with  
this kind of input file:

####
read
    mol pqr G1656D.pqr
end
elec
    mg-para
    async 0
    dime 97 129 129
    pdime 4 3 3
    ofrac 0.1
    cglen 118.0412 92.1046 64.9655
    fglen 89.4360 74.1792 58.2150
    cgcent mol 1
    fgcent mol 1
    mol 1
    lpbe
    bcfl sdh
    pdie 2.0000
    sdie 78.5400
    srfm smol
    chgm spl2
    sdens 10.00
    srad 1.40
    swin 0.30
    temp 298.15
    calcenergy total
    calcforce no
    write pot dx pot
end

quit
###

However, in the output file, I noticed this warning:
###
focusFillBound: WARNING! Unusually large potential values detected on  
the focusing boundary! Convergent not guaranteed!
###

And the resulting dx file contains only 81 x 108 x 129 points (and not  
97 * 129 * 129 as indicated in the input file).
This results in some 'emtpty' space in the complete map !!
I tried sdh and mdh as boundary conditions (same result).
Should I use focus, or map ?


Thanks a lot.

Nicolas Coquelle.


------------------------------------------------------------------------------

_______________________________________________
apbs-users mailing list
apbs-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/apbs-users