Jeremy,
Can you or someone explain in plane english what is going on with ECal
clustering algorithm.
We have the code ready for half a year now, checked with simulations and
documented as a
HPS-NOTE. Now that we have data we seems cannot use that algorithm for
some technical
reasons, or I am wrong. Why the data structure is getting defined now,
how much effort will
be to have Holly's clustering algorithm running as the main and perhaps
the only clustering
code for ECal.
Thanks, Stepan
On 1/5/15 8:23 PM, McCormick, Jeremy I. wrote:
> -----Original Message-----
> From: McCormick, Jeremy I.
> Sent: Monday, January 05, 2015 4:47 PM
> To: 'Holly Vance'
> Cc: 'Kyle McCarty'; Uemura, Sho
> Subject: HPSEcalCluster
>
> Hi,
>
> Please do not add any more usages of HPSEcalCluster or HPSEcalClusterIC to our code base, if at all possible.
>
> The problem here is that this specific type is never persisted. Only the Cluster.class type is known to LCIO.
>
> So Driver code like this will break whenever an LCIO file is read back in for analysis...
>
> event.get(HPSEcalCluster.class, "clusters");
>
> Also, as you probably know, none of the extra information added to HPSEcalCluster is persisted to LCIO.
>
> I am going to change a bunch of files to reference Cluster.class instead of HPSEcalCluster.class in order to try and fix this where I can.
>
> I'm hoping eventually we can completely remove from the code the HPSEcalCluster and HPSEcalClusterIC classes.
>
> In order to this, I'm going to assume that all Cluster objects have as the first CalorimeterHit in their list the seed hit e.g.
>
> aCluster.getCalorimeterHits().get(0);
>
> will be assumed to have the seed.
>
> If this is not the case, then you should change your algorithms so that this is always the case.
>
> Also, we should externalize the position calculations in HPSEcalCluster into a ClusterPropertyCalculator. This does not need to be on a Cluster extension class.
>
> Any questions/concerns here?
>
> --Jeremy
>
> ########################################################################
> Use REPLY-ALL to reply to list
>
> To unsubscribe from the HPS-SOFTWARE list, click the following link:
> https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1
########################################################################
Use REPLY-ALL to reply to list
To unsubscribe from the HPS-SOFTWARE list, click the following link:
https://listserv.slac.stanford.edu/cgi-bin/wa?SUBED1=HPS-SOFTWARE&A=1
|