Skip to Content.
Sympa Menu

opal - Re: [Opal] problem with BINOMIAL distribution mixing up y and t

opal AT lists.psi.ch

Subject: The OPAL Discussion Forum

List archive

Re: [Opal] problem with BINOMIAL distribution mixing up y and t


Chronological Thread 
  • From: Andreas Adelmann <andreas.adelmann AT psi.ch>
  • To: "<suzie.sheehy AT stfc.ac.uk>" <suzie.sheehy AT stfc.ac.uk>
  • Cc: opal AT lists.psi.ch, Yang Jianjun <jianjun.yang AT psi.ch>
  • Subject: Re: [Opal] problem with BINOMIAL distribution mixing up y and t
  • Date: Fri, 4 May 2012 10:18:13 +0200
  • List-archive: <https://lists.web.psi.ch/pipermail/opal/>
  • List-id: The OPAL Discussion Forum <opal.lists.psi.ch>

Dear Suzie in OPAL-cycl the T-direction (wee soon call it Z) is vertical.

This has historical and "comunity" reasons, in the cyclotron world the x-y plane 
is horizontal (mid-plane) and z is vertical. The other strange notation (T) is an
inherited confusion from MAD. 

My plan for the next release of OPAL is, to also accept Z instead of T, however in
OPAL-cycl T (Z) will always be vertical. We are refactoring/writing the manual and 
hopefully describe this in a better way and also add more examples. Sorry for the 
slow progress.

@Jianjun, please correct me if I am wrong :-)

Cheers Andreas 

Hi there,

I am using the pre-compiled binaries installed on our local computing
cluster, but I have also seen this issue when using my latest checkout of
OPAL on my desktop, svn rev 13723.

When using OPAL-Cycl with the DISTRIBUTION command, the manual (1.1.9)
says to define the distribution using sigmas of X, PX, Y, PY and T, PT.
Given the description in the manual, this ought to be horizontal, vertical
and then longitudinal (obvious because T must be time & therefore
longitudinal!)

I was having issues with matching so tried two different binomial
distribution inputs, one with the parameters as I'd expect from the manual
: (which produced sc-ffag-030512-1.h5)

Dist1:DISTRIBUTION, DISTRIBUTION=BINOMIAL,
sigmax=  5.51175734164e-3, sigmapx=1277.28350202, corrx=0.0,mx=1.0,
sigmay=  5.92548981662e-3, sigmapy=1105.14435292, corry=0.0,my=1.0,
sigmat=  10.0e-3, sigmapt=1, corrt=0.0,mt=1.0;


And the other where I copied the y parameters over to t to confirm my
suspicions, which produced sc-ffag-030512-t.h5:

Dist1:DISTRIBUTION, DISTRIBUTION=BINOMIAL,
sigmax=  5.51175734164e-4, sigmapx=127.728350202, corrx=0.0,mx=1.0,
sigmay=  5.92548981662e-4, sigmapy=110.514435292, corry=0.0,my=1.0,
sigmat=  5.92548981662e-4, sigmapt=110.514435292, corrt=0.0,mt=1.0;


As you can see I have changed only the *longitudinal* parameters, and yet
in the screenshots from the output h5 files, the change occurs in z, which
in h5 is the *vertical* direction. There is no issue with h5 here, the
issue is in the binomial distribution command (or at least in the
manual!).


I notice also that there isn't an example of using BINOMIAL in the manual
-- so it's possible I've done something wrong there due to lack of
examples! Could someone please confirm this issue and rectify!?

Thanks,
Suzie


--
Scanned by iCritical.

<ypy-comparison.png><zpzcomparison.png>

------
Dr. sc. math. Andreas (Andy) Adelmann
Staff Scientist
Paul Scherrer Institut WBBA/219 CH-5232 Villigen PSI
Phone Office: xx41 56 310 42 33 Fax: xx41 56 310 31 91
Phone Home: xx41 62 891 91 44
-------------------------------------------------------
Friday: ETH CAB H 8x.x  +41 44 632 xx xx
============================================
The more exotic, the more abstract the knowledge, 
the more profound will be its consequences.
Leon Lederman 
============================================






Archive powered by MHonArc 2.6.19.

Top of Page