Re: [ESPResSo] Testing the analyze function

[Axel Arnold]

On Friday 28 September 2007 14:52, Rich Hanes wrote:
> Good afternoon one and all. I have got espresso working and I ran through
> the first tutorial in the user guide. I can run the simulation no problem
> and get a collection of output files config_0 -> config_19. I wrote the
> analysis script exactly as stated in the user guide (in fact I copied the
> text directly from the user guide and pasted into a new file), but the
> generated rdf.data has all zero's for the value of g(r) for all values of
> r. I have attached all the scripts, a sample config_0 file and the
> generated rdf.data. Is there some other way I can test the analyze
> function, ie with one of the sample scripts to make sure it is working
> properly?
Hi,

unforunately, there are some small bugs in the provided scripts. In 
analysis.tcl, it should be
        set avg_rdf [vecscale [expr 1./$cnt] $avg_rdf]
otherwise, 1/$cnt is done as integer division, which gives 0.

Second, the config-files as we write them currently do not contain the box 
dimensions (only the tcl-variable containing the box length). You can either 
manually set them again in the analyze.tcl by using the same start as in 
tutorial1.tcl, namely
        set n_part 200; set density 0.7
        set box_l [expr pow($n_part/$density,1./3.)]
        setmd box_l $box_l $box_l $box_l
        setmd periodic 1 1 1
or you save the MD-parameters. Then you need to change tutorial1.tcl by 
inserting
          blockfile $f write variable box_l
right before the blockfile $f write particles.

Sorry for the inconvenience, but apparently no one has tried to really code 
the scripts from the introduction before to find these bugs. I thought I 
copied the code correctly from my scripts, but that was apparently a bit 
optimistic.

Many regards,
Axel

-- 
Dr. Axel Arnold
FOM Institute for Atomic and Molecular Physics
Kruislaan 407                                   Phone: +31 20 6081 275
1098 SJ Amsterdam, The Netherlands              E-mail: address@hidden