Re: [fluka-discuss]: Post processing USRBIN file (*.bnn format)

From: <me_at_marychin.org>
Date: Thu, 23 Oct 2014 20:06:43 -0700 (PDT)

Dear Rafiq,

No no no, please don't use FLAIR 50 times to generate 50 dat files. After the 50
runs complete, do FLAIR > Run > Data > Process. This will gather together all
the *_fort.?? files and based on these, produce one merged bnn. Then go to plot,
and plot this merge bnn. You will get a dat file which contains the merged
results of your 50 runs.

:) mary

> On 23 October 2014 at 18:52 Mohammad Islam <rafiqul.islam_at_okstate.edu> wrote:
>
>
> Dear mary,
>
> Thanks for your detail explanation. In my problem i run 50 input files using
> flair simultaneously. I would love to take the .dat file that comes from flair
> but in such case i have to use flair 50 times to generate 50 dat file. Right?
> Thats why i thought to use the available bnn files. Do you know any ways that
> will generate the dat files for all the bnn files at the same time using
> identical flair setup? Thanks for all your help.
>
> Rafiq
>
> On Oct 23, 2014 7:19 PM, " me_at_marychin.org <mailto:me_at_marychin.org> " <
> me_at_marychin.org <mailto:me_at_marychin.org> > wrote:
> > > Dear Rafiq,
> >
> > USRBIN scoring is 3D in general (unless eg you ask for nbin=1 for one of
> > the dimensions). When we do a FLAIR plot for USRBIN it is typically 2D
> > (colormap). dat is the output from FLAIR plot, so it is also typically 2D,
> > unless you specifically ask for 1D projection (line plot). During FLAIR plot
> > for USRBIN, windowing / averaging takes place when we *select* the ranges
> > under 'Projection & Limits'. Note that this *select*ion takes place after
> > the simulation, as opposed to the specification in the USRBIN card before
> > the simulation. This windowing process makes dat a subset of bnn (tons of
> > data, as you say). bnn contains what you asked for in the USRBIN card. dat
> > contains what you ask for in FLAIR plot, and possibly some
> > re-normalisations.
> >
> > If you want a file similar to dat, why don't you just take dat itself, as
> > I mentioned earlier as one of the alternatives. Just make a duplicate of
> > dat, throw off the header lines, then, in the case of 1D, in MATLAB:
> > a = load('rafiqsfile.txt');
> > stairs(a(:,1),a(;,3))
> > --OR--
> > errorbar((a(:,1)+a(:,2))/2,a(:,3),a(:,3).*a(:,4)/100)
> > --OR--
> > plot((a(:,1)+a(:,2))/2,a(:,3))
> > --OR--
> > whatever you prefer.
> >
> > In the case of more dimensions, you'd need a loop to rearrange 'a' into a
> > matrix (in more general terms, 'array'). 2D matrix if 2D. 3D matrix if 3D.
> > If 2D:
> > imagesc(log10(b))
> > If 3D:
> > imagesc(log10(b(:,:,cut_rafiq_wants_to_see)))
> >
> > :) mary
> >
> > > On 23 October 2014 at 13:19 Mohammad Islam < rafiqphy_at_gmail.com
> > > <mailto:rafiqphy_at_gmail.com> > wrote:
> > >
> > >
> > > Dear Mary,
> > >
> > > Thank you very much for your prompt answer. As your suggestion, I
> > > converted using the approach => FLAIR>Run>Files>data then right-click
> > > on the bnn file and select 'To ascii'. But now I see a tons of data
> > > point (although most of them are zero), however if I plot the original
> > > bnn file using flair it gives me a simple *dat file of the whatever
> > > bin width that I had chosen in the input. Can you please tell me how
> > > can I interpret this ascii file. I want a similar .dat file that flair
> > > produces which is just a bragg curve in this case. Attached is one of
> > > the converted ascii file for your comments. Thank you very much.
> > >
> > > On Thu, Oct 23, 2014 at 2:51 PM, me_at_marychin.org
> > > <mailto:me_at_marychin.org> < me_at_marychin.org <mailto:me_at_marychin.org> >
> > > wrote:
> > > > Dear Rafiq,
> > > >
> > > > bnn is binary. If you'd like to do it the brute-force way, use MATLAB
> > > > fread.
> > > >
> > > > The easy way is to convert bnn to ascii, under FLAIR>Run>Files>data
> > > > then
> > > > right-click on the bnn file and select 'To ascii'. Thereafter, use
> > > > MATLAB
> > > > fscanf, fgets, fgetl or ...
> > > >
> > > > An alternative to the 'To ascii' function would be just to take the
> > > > dat
> > > > file, assuming you plotted with FLAIR.
> > > >
> > > > The lazyman's way is: manually edit the ascii file by removing the
> > > > header
> > > > lines. Thereafter, use MATLAB load. This way, you won't even need to
> > > > worry
> > > > about bytes and fields. You won't even need to open and close the
> > > > file.
> > > >
> > > > MATLAB or PYTHON or PERL the algorithm would be similar (with the
> > > > exception
> > > > that MATLAB's load condenses all the steps in to one):
> > > > open the file
> > > > read each line, split according to the spaces or fields or bytes
> > > > arrange the numbers into an array
> > > > close the file
> > > >
> > > > :) mary
> > > >
> > > >> On 23 October 2014 at 11:32 Mohammad Islam <
> > > >> rafiqul.islam_at_okstate.edu <mailto:rafiqul.islam_at_okstate.edu> >
> > > >> wrote:
> > > >>
> > > >>
> > > >> Dear Users,
> > > >>
> > > >> Does anybody know how to read the *.bnn file that comes from USRBIN
> > > >> using MATLAB or any other language. I tried to read it through
> > > >> MATLAB
> > > >> but it did not give me the data I was looking for. I need to do some
> > > >> post processing of the deposited energy which is in number of binary
> > > >> files. FLAIR can read and plot it and also save it as *.dat file but
> > > >> I
> > > >> have so many files that a little bit coding would be helpful. I
> > > >> appreciate any help.
> > > >>
> > > >> Thanks
> > > >> Rafiq
> > > >>
> > >
> > >
> > >
> > > --
> > > Mohammad Rafiqul Islam, PhD
> > >
Received on Fri Oct 24 2014 - 06:30:37 CEST