[Clas12_verystrange] Vertex code (fwd)

[Michael Dugger]

Hi,

Evgeny Golovach sent me instructions regarding the vertex code (see 
bottom of email).

The code is in Java and my analysis is in C++. I have been trying to call 
the Java code from C++ using JNI but when I create the Java virtual 
machine inside C++ I get a core dump. In short, I'm having a heck of a 
time calling Java from C++. I have never done this before :(

This is the code that I'm having problems with:

   JavaVM *jvm;
   JNIEnv *env;
   JavaVMInitArgs vm_args;
   JavaVMOption* options = new JavaVMOption[1];
   sprintf(tmpString,"-Djava.class.path=./");
   options[0].optionString = tmpString;
   vm_args.version = JNI_VERSION_1_6;
   vm_args.nOptions = 1;
   vm_args.options = options;
   vm_args.ignoreUnrecognized = false;
   int rc = JNI_CreateJavaVM(&jvm, (void**)&env, &vm_args);
   //CORE DUMP HERE

Does anyone know what I am doing wrong, or have some example code that 
could help?

Thanks for your time.

Sincerely,
Michael

---------- Forwarded message ----------
Date: Tue, 7 Nov 2017 13:10:45 -0500
From: golovach at jlab.org
To: Michael Dugger <dugger at jlab.org>
Subject: Re: Vertex code

Hi Michael,

Here the vtx.jar file with the vertexing package.
It is not a part of COATJAVA at the moment.
You need to use it to compile and run your code.
You should use it with the reconstructed data files.
(coatjava_4a.8.3)
There is also the vertex.java file.
It is a kind of explanation how to use the code.
The files are in
/lustre/expphy/volatile/clas/clasg11/golovach/VERTEX
(mail system did't allow me send jar's by attachment.)

At the moment vertex reconstruction works only if both
particle tracks are in Forward detector (DC).
It does not work for CVT.
Veronique told me that some kind of "refit of the track finding"
need to be implemented.

Let me know if you have questions/problems.

Evgeny


>
> Evgeny,
>
> Thanks a bunch for working on the tests :)
>
> Take care,
> Michael
>
> On Thu, 2 Nov 2017, golovach at jlab.org wrote:
>
>> Hi Michael,
>>
>> I will send you the code and instructions
>> by Monday. I would like to make quick tests,
>> as I did the tests ~half a year ago and I would like
>> to verify it one more time.
>> Vertex reconstruction code is not a part of
>> COATJAVA for now.
>>
>> Regards,
>>  Evgeny
>>
>>>
>>> Dear Evgeny,
>>>
>>> My name is Michael Dugger and I am working with the Cascade group
>>> (called
>>> the "Very Strange" group).
>>>
>>> Eugene Pasyuk said that you were the person to contact regarding code
>>> that
>>> can be used to determine vertices in reconstructed CLAS12 data.
>>>
>>> As you can imagine, having good vertex information for strangeness
>>> processes is very important. The Very Strange group is excited at the
>>> possibility of gaining a good vertex resolution.
>>>
>>> Any information you can give me as to how to use your code to analyze
>>> the
>>> vertex positions for reconstructed CLAS12 data would be greatly
>>> appreciated.
>>>
>>> Thanks for your time.
>>>
>>> Sincerely,
>>> Michael
>>>
>>
>>
>>
>