Well, let me rephrase - the values all have to be the same type if you
choose to read all columns in a similar way. If you know in advance its
always the value associated with key 'b' that's a bag, why don't you cast
that single value?
B = foreach A generate (bag{})document#'b' as b;
On Thu, Apr 18, 2013 at 7:43 AM, Jerry Lam <[email protected]> wrote:
> Hi Prashant:
>
> I read about the map data type in the book "Programming Pig", it says:
> "... By default there is no requirement that all values in a map must be of
> the same type. It is legitimate to have a map with two keys name and age,
> where the value for name is a chararray and the value for age is an int.
> Beginning in Pig 0.9, a map can declare its values to all be of the same
> type... "
>
> I agree that all values in the map can be of the same type but this is not
> required in pig.
>
> Best Regards,
>
> Jerry
>
>
> On Thu, Apr 18, 2013 at 10:37 AM, Jerry Lam <[email protected]> wrote:
>
> > Hi Rusian:
> >
> > I used PigStorage to store the data that is originally using Pig data
> > type. It is strange (or a bug in Pig) that I cannot read the data using
> > PigStorage that have been stored using PigStorage, isn't it?
> >
> > Best Regards,
> >
> > Jerry
> >
> >
> >
> > On Wed, Apr 17, 2013 at 10:52 PM, Ruslan Al-Fakikh <[email protected]
> >wrote:
> >
> >> The output:
> >> ({ ([c#11,d#22]),([c#33,d#44]) })
> >> ()
> >> looks weird.
> >>
> >> Jerry, maybe the problem is in using PigStorage. As its javadoc says:
> >>
> >> A load function that parses a line of input into fields using a
> character
> >> delimiter
> >>
> >> So I guess this is just for simple csv lines.
> >> But you are trying to load a complicated Map structure as it was
> formatted
> >> by previous storing.
> >> Probably you'll need to write your own Loader for this. Another hint:
> >> using
> >> the -schema paramenter to PigStorage, but I am not sure it can help:(
> >>
> >> Ruslan
> >>
> >>
> >> On Wed, Apr 17, 2013 at 11:48 PM, Jerry Lam <[email protected]>
> wrote:
> >>
> >> > Hi Rusian:
> >> >
> >> > I did a describe B followed by a dump B, the output is:
> >> > B: {b: {()}}
> >> >
> >> > ({ ([c#11,d#22]),([c#33,d#44]) })
> >> > ()
> >> >
> >> > but when I executed
> >> >
> >> > C = foreach B generate flatten(b);
> >> >
> >> > dump C;
> >> >
> >> > I got the exception again...
> >> >
> >> > 2013-04-17 15:47:39,933 [Thread-26] WARN
> >> > org.apache.hadoop.mapred.LocalJobRunner - job_local_0002
> >> > java.lang.Exception: java.lang.ClassCastException:
> >> > org.apache.pig.data.DataByteArray cannot be cast to
> >> > org.apache.pig.data.DataBag
> >> > at
> >> org.apache.hadoop.mapred.LocalJobRunner$Job.run(LocalJobRunner.java:400)
> >> > Caused by: java.lang.ClassCastException:
> >> org.apache.pig.data.DataByteArray
> >> > cannot be cast to org.apache.pig.data.DataBag
> >> > at
> >> >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.physicalLayer.expressionOperators.POProject.processInputBag(POProject.java:586)
> >> > at
> >> >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.physicalLayer.expressionOperators.POProject.getNext(POProject.java:250)
> >> > at
> >> >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.physicalLayer.PhysicalOperator.getNext(PhysicalOperator.java:334)
> >> > at
> >> >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.physicalLayer.relationalOperators.POForEach.processPlan(POForEach.java:372)
> >> > at
> >> >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.physicalLayer.relationalOperators.POForEach.getNext(POForEach.java:297)
> >> > at
> >> >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.mapReduceLayer.PigGenericMapBase.runPipeline(PigGenericMapBase.java:283)
> >> > at
> >> >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.mapReduceLayer.PigGenericMapBase.map(PigGenericMapBase.java:278)
> >> > at
> >> >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.mapReduceLayer.PigGenericMapBase.map(PigGenericMapBase.java:64)
> >> > at org.apache.hadoop.mapreduce.Mapper.run(Mapper.java:144)
> >> > at org.apache.hadoop.mapred.MapTask.runNewMapper(MapTask.java:725)
> >> > at org.apache.hadoop.mapred.MapTask.run(MapTask.java:332)
> >> > at
> >> >
> >> >
> >>
> org.apache.hadoop.mapred.LocalJobRunner$Job$MapTaskRunnable.run(LocalJobRunner.java:232)
> >> > at
> >> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:439)
> >> > at java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303)
> >> > at java.util.concurrent.FutureTask.run(FutureTask.java:138)
> >> > at
> >> >
> >> >
> >>
> java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:895)
> >> > at
> >> >
> >> >
> >>
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:918)
> >> > at java.lang.Thread.run(Thread.java:680)
> >> >
> >> >
> >> > Best Regards,
> >> >
> >> > Jerry
> >> >
> >> >
> >> > On Wed, Apr 17, 2013 at 3:26 PM, Ruslan Al-Fakikh <
> [email protected]
> >> > >wrote:
> >> >
> >> > > I think that before doing the FLATTEN, you should be 100% sure that
> >> your
> >> > > cast worked properly. Can you first DESCRIBE B and then DUMP B right
> >> > away?
> >> > > Or probably it just can't be cast in this way. Honestly I don't know
> >> > > exactly how it works, but here:
> >> > > http://pig.apache.org/docs/r0.10.0/basic.html#cast
> >> > > I see that casting from a map to a bag should produce an error.
> >> > > Hope that helps.
> >> > >
> >> > >
> >> > > On Wed, Apr 17, 2013 at 9:38 PM, Jerry Lam <[email protected]>
> >> wrote:
> >> > >
> >> > > > Hi Rusian:
> >> > > >
> >> > > > Thanks for your help. I really appreciate it. It really puzzled
> me.
> >> > > >
> >> > > > I did a "describe B", the output is "B: {b: bytearray}".
> >> > > >
> >> > > > I then tried to cast it as suggested, I got:
> >> > > > B = foreach A generate document#'b' as b:{};
> >> > > > describe B;
> >> > > > B: {b: {()}}
> >> > > >
> >> > > > Then I proceed with:
> >> > > > C = foreach B generate flatten(b);
> >> > > >
> >> > > > I got:
> >> > > > 2013-04-17 13:38:04,601 [Thread-16] WARN
> >> > > > org.apache.hadoop.mapred.LocalJobRunner - job_local_0002
> >> > > > java.lang.Exception: java.lang.ClassCastException:
> >> > > > org.apache.pig.data.DataByteArray cannot be cast to
> >> > > > org.apache.pig.data.DataBag
> >> > > > at
> >> > >
> >> org.apache.hadoop.mapred.LocalJobRunner$Job.run(LocalJobRunner.java:400)
> >> > > > Caused by: java.lang.ClassCastException:
> >> > > org.apache.pig.data.DataByteArray
> >> > > > cannot be cast to org.apache.pig.data.DataBag
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.physicalLayer.expressionOperators.POProject.processInputBag(POProject.java:586)
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.physicalLayer.expressionOperators.POProject.getNext(POProject.java:250)
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.physicalLayer.PhysicalOperator.getNext(PhysicalOperator.java:334)
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.physicalLayer.relationalOperators.POForEach.processPlan(POForEach.java:372)
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.physicalLayer.relationalOperators.POForEach.getNext(POForEach.java:297)
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.mapReduceLayer.PigGenericMapBase.runPipeline(PigGenericMapBase.java:283)
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.mapReduceLayer.PigGenericMapBase.map(PigGenericMapBase.java:278)
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> org.apache.pig.backend.hadoop.executionengine.mapReduceLayer.PigGenericMapBase.map(PigGenericMapBase.java:64)
> >> > > > at org.apache.hadoop.mapreduce.Mapper.run(Mapper.java:144)
> >> > > > at org.apache.hadoop.mapred.MapTask.runNewMapper(MapTask.java:725)
> >> > > > at org.apache.hadoop.mapred.MapTask.run(MapTask.java:332)
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> org.apache.hadoop.mapred.LocalJobRunner$Job$MapTaskRunnable.run(LocalJobRunner.java:232)
> >> > > > at
> >> > >
> >> java.util.concurrent.Executors$RunnableAdapter.call(Executors.java:439)
> >> > > > at
> >> java.util.concurrent.FutureTask$Sync.innerRun(FutureTask.java:303)
> >> > > > at java.util.concurrent.FutureTask.run(FutureTask.java:138)
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> java.util.concurrent.ThreadPoolExecutor$Worker.runTask(ThreadPoolExecutor.java:895)
> >> > > > at
> >> > > >
> >> > > >
> >> > >
> >> >
> >>
> java.util.concurrent.ThreadPoolExecutor$Worker.run(ThreadPoolExecutor.java:918)
> >> > > > at java.lang.Thread.run(Thread.java:680)
> >> > > >
> >> > > > Best Regards,
> >> > > >
> >> > > > Jerry
> >> > > >
> >> > > >
> >> > > > On Wed, Apr 17, 2013 at 1:24 PM, Ruslan Al-Fakikh <
> >> > [email protected]
> >> > > > >wrote:
> >> > > >
> >> > > > > Hey, and as for converting a map of tuples, probably i got you
> >> wrong.
> >> > > If
> >> > > > > you can get to every value manually withing FOREACH then I see
> no
> >> > > problem
> >> > > > > in doing so.
> >> > > > >
> >> > > > >
> >> > > > > On Wed, Apr 17, 2013 at 9:22 PM, Ruslan Al-Fakikh <
> >> > > [email protected]
> >> > > > > >wrote:
> >> > > > >
> >> > > > > > I am not sure whether you can convert a map to a tuple.
> >> > > > > > But I am curious about one thing:
> >> > > > > > your are trying to use 'b' as a Bag, right? Because FLATTEN
> >> needs
> >> > it
> >> > > to
> >> > > > > be
> >> > > > > > a Bag I guess:
> >> > > > > > http://pig.apache.org/docs/r0.10.0/basic.html#flatten
> >> > > > > > But it seems that Pig thinks that b is a byte array:
> >> > > > > > java.lang.ClassCastException:
> org.apache.pig.data.DataByteArray
> >> > > cannot
> >> > > > be
> >> > > > > > cast to org.apache.pig.data.DataBag
> >> > > > > > Can you do this?:
> >> > > > > > DESCRIBE B
> >> > > > > >
> >> > > > > > I suppose it can look like a Bag in the output of DUMP, but I
> >> think
> >> > > Pig
> >> > > > > > doesn't know it is a Bag, maybe you'll need some kind of
> >> explicit
> >> > > cast?
> >> > > > > >
> >> > > > > >
> >> > > > > > On Wed, Apr 17, 2013 at 9:11 PM, Jerry Lam <
> >> [email protected]>
> >> > > > wrote:
> >> > > > > >
> >> > > > > >> Hi Rusian,
> >> > > > > >>
> >> > > > > >> I tried to debug each step already but no luck.
> >> > > > > >> I did a dump (dump B;) after B = foreach A generate
> >> document#'b'
> >> > as
> >> > > b;
> >> > > > > >> I got {([c#11,d#22]),([c#33,d#44])}
> >> > > > > >> but it fails when I did C = foreach B generate flatten(b);
> >> > > > > >>
> >> > > > > >> I don't have controls over the input. It is passed as Map of
> >> > Maps. I
> >> > > > > guess
> >> > > > > >> it makes lookup easier using a map with keys.
> >> > > > > >>
> >> > > > > >> Can I convert map to tuple?
> >> > > > > >>
> >> > > > > >> Best Regards,
> >> > > > > >>
> >> > > > > >> Jerry
> >> > > > > >>
> >> > > > > >>
> >> > > > > >>
> >> > > > > >> On Wed, Apr 17, 2013 at 11:57 AM, Ruslan Al-Fakikh <
> >> > > > > [email protected]
> >> > > > > >> >wrote:
> >> > > > > >>
> >> > > > > >> > Hi Jerry,
> >> > > > > >> >
> >> > > > > >> > I would recommend to debug the issue step by step. Just
> after
> >> > this
> >> > > > > line:
> >> > > > > >> > A = load 'data.txt' as document:[];
> >> > > > > >> > and then right after that:
> >> > > > > >> > DESCRIBE A;
> >> > > > > >> > DUMP A;
> >> > > > > >> > and so on...
> >> > > > > >> >
> >> > > > > >> > To be honest I haven't used maps that much. Just curious,
> why
> >> > did
> >> > > > you
> >> > > > > >> > choose to use them? You can also use regular tuples for
> >> storing
> >> > > the
> >> > > > > >> > relations. Also you can store the tuples with a schema
> file.
> >> > > > > >> >
> >> > > > > >> > Ruslan
> >> > > > > >> >
> >> > > > > >> >
> >> > > > > >> > On Wed, Apr 17, 2013 at 5:28 AM, Jerry Lam <
> >> > [email protected]>
> >> > > > > >> wrote:
> >> > > > > >> >
> >> > > > > >> > > Hi pig users,
> >> > > > > >> > >
> >> > > > > >> > > I tried to load data using PigStorage that was previously
> >> > stored
> >> > > > > using
> >> > > > > >> > > PigStorage but it failed.
> >> > > > > >> > >
> >> > > > > >> > > Each line looks like this in the data file that is
> >> generated
> >> > by
> >> > > > > >> > PigStorage:
> >> > > > > >> > > [a#hello,b#{([c#11,d#22]),([c#33,d#44])}]
> >> > > > > >> > >
> >> > > > > >> > > I did the following:
> >> > > > > >> > > A = load 'data.txt' as document:[];
> >> > > > > >> > > B = foreach A generate document#'b' as b;
> >> > > > > >> > > C = foreach B generate flatten(b);
> >> > > > > >> > > dump C;
> >> > > > > >> > >
> >> > > > > >> > > I expect to see the following output:
> >> > > > > >> > > ([c#11,d#22])
> >> > > > > >> > > ([c#33,d#44])
> >> > > > > >> > >
> >> > > > > >> > > Instead, I got:
> >> > > > > >> > > java.lang.ClassCastException:
> >> > org.apache.pig.data.DataByteArray
> >> > > > > >> cannot be
> >> > > > > >> > > cast to org.apache.pig.data.DataBag
> >> > > > > >> > >
> >> > > > > >> > > Anyone encounters this problem before? How can I read the
> >> data
> >> > > > back?
> >> > > > > >> > >
> >> > > > > >> > > Thanks,
> >> > > > > >> > >
> >> > > > > >> > > Jerry
> >> > > > > >> > >
> >> > > > > >> >
> >> > > > > >>
> >> > > > > >
> >> > > > > >
> >> > > > >
> >> > > >
> >> > >
> >> >
> >>
> >
> >
>
