Thanks.  Indeed there is not, so that is good news for me.



On Tue, Mar 16, 2010 at 9:18 PM, Paul Yang <[email protected]> wrote:

>  I forgot to mention, if there is no WHERE clause in the query, then the
> query will work fine.
>
>
>
> *From:* Tim Robertson [mailto:[email protected]]
> *Sent:* Tuesday, March 16, 2010 12:20 PM
>
> *To:* [email protected]
> *Subject:* Re: Expanding comma separated values in a column
>
>
>
> Ok, thanks Paul.
>
>
>
> Well, it looked like it worked anyway, but I will rerun... I got the
> expected number of records on the output, but have not yet loaded back into
> Mysql or put the  the web app on it to see how the data looks.
>
>
>
> I'm really very impressed with Hive.  Looking forward to hooking that up to
> HBase and showing what we can now achieve over our struggling Mysql
> environment.
>
>
>
> Cheers,
>
> Tim
>
>
>
>
>
>
>
> On Tue, Mar 16, 2010 at 8:07 PM, Paul Yang <[email protected]> wrote:
>
> Hey Tim,
>
>
>
> Just FYI, there is a bug with lateral views and the predicate push down
> that will be patched very soon. Until then, you might need to use:
>
>
>
> set hive.optimize.ppd = false;
>
>
>
> Cheers,
>
> Paul
>
>
>
>
>
> *From:* Tim Robertson [mailto:[email protected]]
> *Sent:* Tuesday, March 16, 2010 6:30 AM
> *To:* [email protected]
> *Subject:* Re: Expanding comma separated values in a column
>
>
>
> Perfect.  Thanks
>
>
>
> On Tue, Mar 16, 2010 at 11:36 AM, Sonal Goyal <[email protected]>
> wrote:
>
> Hi Tim,
>
> You can use the explode UDTF. More here:
>
> http://wiki.apache.org/hadoop/Hive/LanguageManual/LateralView
>
> HTH
> Thanks and Regards,
> Sonal
>
>
>
> On Tue, Mar 16, 2010 at 3:32 PM, Tim Robertson <[email protected]>
> wrote:
>
> Hi all,
>
>
>
> I have a table of 2 columns of strings, with example row as:
>
>
>
> Col1      Col2
>
> 123      23,34,45,67... up to around 1 million
>
>
>
> I'd like to expand the comma separated values to a new taller KVP table:
>
>
>
> Col1  Col2
>
> 123    23
>
> 123    34
>
> 123    45
>
> 123    67
>
> 123    .... potentially 1,000,000 rows generated
>
>
>
> Can someone please point me in the right direction?
>
>
>
> Thanks
>
> Tim
>
>
>
>
>
>
>
>
>
>
>

Reply via email to