[Halld-offline] indiosyncrasies of FFREAD

[Richard Jones]

David,

So THAT'S what busted it!

Ok, thanks, I thought I had forgotten the default behavior, but someone was fiddling.  In the future, if you need a change in the fundamental behavior (structure of the input file) of an application, could I ask that you contact the principal author/maintainer?   Too many cooks ...

One day, the DEBUG command in my control.in suddenly stopped working, with no explanation.  Eventually I tried reducing it to DEBU, and bingo.

-Richard J.


On 3/21/2011 3:39 PM, David Lawrence wrote:
> Hi Richard,
>
>      I thought the call to FFSET('SIZE', 16) that Sascha put in made the max size 16 characters opening up the naming space from the 4 character limit a bit. Can anyone else confirm this?
>
> Regards,
> -Dave
>
> On 3/21/11 3:18 PM, Richard Jones wrote:
> Kei,
>
> Another one of those follies from FFREAD that I recently re-learned is that it matters how many letters you put in the leading word in the lines of control.in.  If the full command name is more than 4 characters then any string of 4 or more characters that coincides with the name works.  For example, any of the following cards will work for the INFILE command:
>
>
>   *   INFI   'rhop.hddm'<--<<  ok
>   *   INFIL 'rhop.hddm'<--<<  ok
>   *   INFILE 'rhop.hddm'<--<<  ok
>
> HOWEVER, if the specification is a shortened form, then spelling out the complete word fails.  For example, the command for debug output is spelled DEBU (without the G).   The first one below works, while the second one is silently ignored.
>
>   *   DEBU   1 10 1000<--<<  ok
>   *   DEBUG 1 10 1000<--<<  bad form
>
> Just thought you might like to know.
>
> -Richard J.
>
>
>
> On 3/21/2011 2:43 PM, Kei Moriya wrote:
>
> Dear offliners,
>
> I wanted to let you know that there is a little bug
> when generating particles with a spread in
> momentum and angle with the particle gun.
>
> When generating events with the particle gun,
> I specify the central value of p, theta, and phi,
> along with the full width of each value.
> What I have noticed is that if I specify for
> example
> c   particle  momentum  theta  phi  delta_momentum delta_theta delta_phi
> KINE   114      5.0       30.   0.      0.5              30        360.
>
> which means generate protons(offset=100+PID=14) with
> p = 5.0 +/- 5
> theta = 30 +/- 15
> phi = 0 +/- 180.
>
> What I have noticed is that the decimal point for
> the momentum values are essential, since without
> the decimal point, the spread becomes 0!
> I assume this is a simple bug of taking the value
> as an int instead of a float.
>
> The plot attachd show the difference when
> a "." is added to delta_momentum so that the
> top plot (with decimal point) has a full width
> of 5 (2.5<p<7.5), while the bottom (without decimal
> point) has no width.
>
> I assume the place to correct this is within
> gukine.F in sim-recon/sim-recon/src/programs/Simulations/HDGeant/
> but I'll let the experts fix this so I don't
> break anything.
>
> One thing that comes as a consequence of this is that
> the plots I showed last week had theta exactly
> at 0.25,0.75,1.25, ... degrees, and NO spread.
> I'm not sure how many other people are using the
> particle gun, but you may want to check that you
> are actually getting a spread in your values.
>
>          Kei
>
>
>
>
>
>
>
> _______________________________________________
> Halld-offline mailing list
> Halld-offline at jlab.org<mailto:Halld-offline at jlab.org>
> https://mailman.jlab.org/mailman/listinfo/halld-offline


-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 4092 bytes
Desc: S/MIME Cryptographic Signature
Url : https://mailman.jlab.org/pipermail/halld-offline/attachments/20110321/eca47db9/attachment.bin