Hi, Kishore,

Thanks for the confirmation.  I was trying to directly place the result
into the Java int array without going through a C array, but wasn't able to
do it the right way.

If anyone knows how to avoid this extra copying, please let us know.

Thanks.

John



On Thu, Apr 4, 2013 at 9:18 AM, kishore g <[email protected]> wrote:

> Thanks John, this worked.
>
>
> On Thu, Apr 4, 2013 at 6:22 AM, K. John Wu <[email protected]> wrote:
>
>> HI, Kishore,
>>
>> I found a problem with the use of java array in the JNI function for
>> get_result_row_ids.  This should have been fixed in SVN 622.  Please
>> give it a try.
>>
>> Thanks.
>>
>> John
>>
>>
>>
>> On 4/3/13 9:47 AM, kishore g wrote:
>> > Thanks John, I tried that and looks like jvm crashed. I tried
>> > debugging the core file but could not get much info from the dump.
>> >
>> > I just added the following to milky.java processQuery method
>> >
>> > int[] rowIds= fb.get_result_row_ids(h);
>> > if( rowIds!=null){
>> >  System.out.println("START row ids");
>> >
>> >  for(int i=0;i<rowIds.length;i++){
>> >   System.out.println(rowIds[i]);
>> >  }
>> >  System.out.println("END row ids");
>> > }
>> >
>> >
>> >
>> > On Tue, Apr 2, 2013 at 12:03 PM, K. John Wu <[email protected]
>> > <mailto:[email protected]>> wrote:
>> >
>> >     Hi, Kishore,
>> >
>> >     Please check out SVN revision 621.  The new function is
>> >     FastBit::get_result_row_ids.
>> >
>> >     Let me know if it works for you or not.
>> >
>> >     John
>> >
>> >
>> >     On 3/30/13 10:46 PM, kishore g wrote:
>> >     > Thanks John, that was exactly what I was looking for. Any chance
>> of
>> >     > exposing this in the jni wrapper?
>> >     >
>> >     > thanks,
>> >     > Kishore G
>> >     >
>> >     >
>> >     > On Sat, Mar 30, 2013 at 8:16 PM, K. John Wu <[email protected]
>> >     <mailto:[email protected]>
>> >     > <mailto:[email protected] <mailto:[email protected]>>> wrote:
>> >     >
>> >     >     Hi, Kishore,
>> >     >
>> >     >     If you use the ibis::query class, there is a function
>> >     getHitRows that
>> >     >     will return the row ids.
>> >     >
>> >     >     If you want to get at the bit vector directly, you can use the
>> >     >     function getHitVector.
>> >     >
>> >     >     John
>> >     >
>> >     >
>> >     >     On 3/30/13 10:09 AM, kishore g wrote:
>> >     >     > Hi,
>> >     >     >
>> >     >     > Is there a efficient way to retrieve the bitvector after
>> >     >     applying the
>> >     >     > predicates.
>> >     >     >
>> >     >     > One way would be to add additional row_id column to the
>> >     data and
>> >     >     then
>> >     >     > use select row_id where col1=v1 col2=v2 and iterate over
>> >     the row_id
>> >     >     > but this is inefficient because it accesses the row_id col
>> on
>> >     >     disk to
>> >     >     > retrieve the values. Instead we want to simply get the bit
>> >     positions
>> >     >     > that are set to 1 after evaluating the predicates.
>> >     >     >
>> >     >     > Are there api's that allow me to get the resultant bit
>> >     vector after
>> >     >     > evaluating  the predicates?
>> >     >     >
>> >     >     > thanks,
>> >     >     > Kishore G
>> >     >     >
>> >     >     >
>> >     >     > _______________________________________________
>> >     >     > FastBit-users mailing list
>> >     >     > [email protected]
>> >     <mailto:[email protected]>
>> >     <mailto:[email protected]
>> >     <mailto:[email protected]>>
>> >     >     >
>> https://hpcrdm.lbl.gov/cgi-bin/mailman/listinfo/fastbit-users
>> >     >     >
>> >     >
>> >     >
>> >
>> >
>>
>
>
_______________________________________________
FastBit-users mailing list
[email protected]
https://hpcrdm.lbl.gov/cgi-bin/mailman/listinfo/fastbit-users

Reply via email to