H5Part development and discussion

[Kurt Stockinger <KStockinger AT lbl.gov>]

John Shalf wrote:
>
> Hi Kurt,
> I think SetView is inclusive of the starting particle address but
> exclusive of the ending address.  I can't remember the rationale for
> that, but I think that is the case.
Hi John,

You are right on the behavior of SetView. The problem seems with the
return value of H5GetNumOfParticles:

H5Part.c:
h5part_int64_t
_H5Part_get_num_particles (
..
    if ( H5PartHasView ( f ) ) {
        nparticles = H5Sget_select_npoints ( space_id );
    }

nparticles is off by 1.

Kurt
>
> -john
>
> On Dec 1, 2006, at 2:23 PM, Kurt Stockinger wrote:
>
>> Hi,
>>
>> I just noticed that H5PartGetNumParticles() seems to return the wrong
>> number of particles. It's always off by one, i.e. n-1 rather than n.
>> I've tested the program H5PartTest with 4 processes and 5 particles
>> each.
>>
>> Thanks,
>> Kurt
>>
>> trunk/test> mpirun -np 4 ./H5PartTest
>> Proc[0] Writing timestep 0 file=67108864
>> ..
>> Proc[3] Writing timestep 4 file=67108864
>> AllDone p[1]
>> ..
>> After SetView(0,4): steps= 5  datasets= 7 particles= 4
>> After SetView(15,19): steps= 5  datasets= 7 particles= 4
>> nParticles: 4
>> nParticles: 4
>> After SetView(10,14): steps= 5  datasets= 7 particles= 4
>> nParticles: 4
>> After SetView(5,9): steps= 5  datasets= 7 particles= 4
>> nParticles: 4
>> proc[2]:  done
>> proc[1]:  done
>> proc[3]:  done
>> proc[0]:  done
>>
>>
>> trunk/test> h5dump parttest.h5
>> HDF5 "parttest.h5" {
>> GROUP "/" {
>>    GROUP "Step#0" {
>>       DATASET "id" {
>>          DATATYPE  H5T_STD_I64LE
>>          DATASPACE  SIMPLE { ( 20 ) / ( 20 ) }
>>          DATA {
>>          (0): 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15,
>> 16, 17,
>>          (18): 18, 19
>>          }
>>       }
>>       DATASET "px" {
>>          DATATYPE  H5T_IEEE_F64LE
>>          DATASPACE  SIMPLE { ( 20 ) / ( 20 ) }
>>          DATA {
>>          (0): 0, 1, 2, 3, 4, 10, 11, 12, 13, 14, 20, 21, 22, 23, 24,
>> 30, 31,
>>          (17): 32, 33, 34
>>          }
>>       }
>>
>>
>> --Kurt Stockinger
>> Computational Research Division
>> Lawrence Berkeley National Laboratory
>> Mail Stop 50B-3238, 1 Cyclotron Road
>> Berkeley, California 94720, USA
>>
>> Tel: +1 (510) 486 5208, Fax: +1 (510) 486 4004
>> email: KStockinger AT lbl.gov
>> http://sdm.lbl.gov/kurts/
>>
>> _______________________________________________
>> H5Part mailing list
>> H5Part AT lists.psi.ch
>> https://lists.web.psi.ch/mailman/listinfo/h5part
>


-- 
Kurt Stockinger
Computational Research Division
Lawrence Berkeley National Laboratory
Mail Stop 50B-3238, 1 Cyclotron Road
Berkeley, California 94720, USA

Tel: +1 (510) 486 5208, Fax: +1 (510) 486 4004
email: KStockinger AT lbl.gov
http://sdm.lbl.gov/kurts/