Re: Weird results with Spark SQL Outer joins
gt;> d.ad) WHERE s.date >= '2016-01-03' AND d.date >= > '2016-01-03'").count() > >>>> res9: Long = 23809 > >>>> > >>>> > >>>> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , > d.account > >>>> AS d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , > >>>> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s FULL OUTER JOIN > >>>> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND > s.ad = > >>>> d.ad) WHERE s.date >= '2016-01-03' AND d.date >= > '2016-01-03'").count() > >>>> res10: Long = 23809 > >>>> > >>>> > >>>> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , > d.account > >>>> AS d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , > >>>> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s LEFT OUTER JOIN > >>>> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND > s.ad = > >>>> d.ad) WHERE s.date >= '2016-01-03' AND d.date >= > '2016-01-03'").count() > >>>> res11: Long = 23809 > >>>> > >>>> > >>>> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , > d.account > >>>> AS d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , > >>>> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s RIGHT OUTER JOIN > >>>> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND > s.ad = > >>>> d.ad) WHERE s.date >= '2016-01-03' AND d.date >= > '2016-01-03'").count() > >>>> res12: Long = 23809 > >>>> > >>>> > >>>> > >>>> From my results above, we notice that the counts of distinct values > based > >>>> on the join criteria and filter criteria for each individual table is > >>>> located at res6 and res7. My question is why is the outer join > producing > >>>> less rows than the smallest table; if there are no matches it should > still > >>>> bring in that row as part of the outer join. For the full and right > outer > >>>> join I am expecting to see a minimum of res6 rows, but I get less, is > there > >>>> something specific that I am missing here? I am expecting that the > full > >>>> outer join would give me the union of the two table sets so I am > expecting > >>>> at least 42533 rows not 23809. > >>>> > >>>> > >>>> Gourav, > >>>> > >>>> I just ran this result set on a new session with slightly newer > data... > >>>> still seeing those results. > >>>> > >>>> > >>>> > >>>> Thanks, > >>>> > >>>> KP > >>>> > >>>> > >>>> On Mon, May 2, 2016 at 11:16 PM, Davies Liu > >>>> wrote: > >>>>> > >>>>> as @Gourav said, all the join with different join type show the same > >>>>> results, > >>>>> which meant that all the rows from left could match at least one row > >>>>> from right, > >>>>> all the rows from right could match at least one row from left, even > >>>>> the number of row from left does not equal that of right. > >>>>> > >>>>> This is correct result. > >>>>> > >>>>> On Mon, May 2, 2016 at 2:13 PM, Kevin Peng wrote: > >>>>>> Yong, > >>>>>> > >>>>>> Sorry, let explain my deduction; it is going be difficult to get a > >>>>>> sample > >>>>>> data out since the dataset I am using is proprietary. > >>>>>> > >>>>>> From the above set queries (ones mentioned in above comments) both > >>>>>> inner and > >>>>>> outer join are producing the same counts. They are basically > pulling > >>>>>> out > >>>>>> selected columns from the query, but there is no roll up happening > or > >>>>>> anything that would possible make it suspicious that there is any > >>>>>> difference > >>>>>> besides the type of joins. The tables are matched based on three >
Re: Weird results with Spark SQL Outer joins
ate >= '2016-01-03' AND d.date >= '2016-01-03'").count() >>>> res11: Long = 23809 >>>> >>>> >>>> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account >>>> AS d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >>>> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s RIGHT OUTER JOIN >>>> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND s.ad >>>> = >>>> d.ad) WHERE s.date >= '2016-01-03' AND d.date >= '2016-01-03'").count() >>>> res12: Long = 23809 >>>> >>>> >>>> >>>> From my results above, we notice that the counts of distinct values based >>>> on the join criteria and filter criteria for each individual table is >>>> located at res6 and res7. My question is why is the outer join producing >>>> less rows than the smallest table; if there are no matches it should still >>>> bring in that row as part of the outer join. For the full and right outer >>>> join I am expecting to see a minimum of res6 rows, but I get less, is there >>>> something specific that I am missing here? I am expecting that the full >>>> outer join would give me the union of the two table sets so I am expecting >>>> at least 42533 rows not 23809. >>>> >>>> >>>> Gourav, >>>> >>>> I just ran this result set on a new session with slightly newer data... >>>> still seeing those results. >>>> >>>> >>>> >>>> Thanks, >>>> >>>> KP >>>> >>>> >>>> On Mon, May 2, 2016 at 11:16 PM, Davies Liu >>>> wrote: >>>>> >>>>> as @Gourav said, all the join with different join type show the same >>>>> results, >>>>> which meant that all the rows from left could match at least one row >>>>> from right, >>>>> all the rows from right could match at least one row from left, even >>>>> the number of row from left does not equal that of right. >>>>> >>>>> This is correct result. >>>>> >>>>> On Mon, May 2, 2016 at 2:13 PM, Kevin Peng wrote: >>>>>> Yong, >>>>>> >>>>>> Sorry, let explain my deduction; it is going be difficult to get a >>>>>> sample >>>>>> data out since the dataset I am using is proprietary. >>>>>> >>>>>> From the above set queries (ones mentioned in above comments) both >>>>>> inner and >>>>>> outer join are producing the same counts. They are basically pulling >>>>>> out >>>>>> selected columns from the query, but there is no roll up happening or >>>>>> anything that would possible make it suspicious that there is any >>>>>> difference >>>>>> besides the type of joins. The tables are matched based on three keys >>>>>> that >>>>>> are present in both tables (ad, account, and date), on top of this >>>>>> they are >>>>>> filtered by date being above 2016-01-03. Since all the joins are >>>>>> producing >>>>>> the same counts, the natural suspicions is that the tables are >>>>>> identical, >>>>>> but I when I run the following two queries: >>>>>> >>>>>> scala> sqlContext.sql("select * from swig_pin_promo_lt where date >>>>>>> ='2016-01-03'").count >>>>>> >>>>>> res14: Long = 34158 >>>>>> >>>>>> scala> sqlContext.sql("select * from dps_pin_promo_lt where date >>>>>>> ='2016-01-03'").count >>>>>> >>>>>> res15: Long = 42693 >>>>>> >>>>>> >>>>>> The above two queries filter out the data based on date used by the >>>>>> joins of >>>>>> 2016-01-03 and you can see the row count between the two tables are >>>>>> different, which is why I am suspecting something is wrong with the >>>>>> outer >>>>>> joins in spark sql, because in this situation the right and outer >>>>>> joins may >>>>>> produce the same results, b
Re: Weird results with Spark SQL Outer joins
e tables are matched based on three > keys > >>>> > that > >>>> > are present in both tables (ad, account, and date), on top of this > >>>> > they are > >>>> > filtered by date being above 2016-01-03. Since all the joins are > >>>> > producing > >>>> > the same counts, the natural suspicions is that the tables are > >>>> > identical, > >>>> > but I when I run the following two queries: > >>>> > > >>>> > scala> sqlContext.sql("select * from swig_pin_promo_lt where date > >>>> >>='2016-01-03'").count > >>>> > > >>>> > res14: Long = 34158 > >>>> > > >>>> > scala> sqlContext.sql("select * from dps_pin_promo_lt where date > >>>> >>='2016-01-03'").count > >>>> > > >>>> > res15: Long = 42693 > >>>> > > >>>> > > >>>> > The above two queries filter out the data based on date used by the > >>>> > joins of > >>>> > 2016-01-03 and you can see the row count between the two tables are > >>>> > different, which is why I am suspecting something is wrong with the > >>>> > outer > >>>> > joins in spark sql, because in this situation the right and outer > >>>> > joins may > >>>> > produce the same results, but it should not be equal to the left > join > >>>> > and > >>>> > definitely not the inner join; unless I am missing something. > >>>> > > >>>> > > >>>> > Side note: In my haste response above I posted the wrong counts for > >>>> > dps.count, the real value is res16: Long = 42694 > >>>> > > >>>> > > >>>> > Thanks, > >>>> > > >>>> > > >>>> > KP > >>>> > > >>>> > > >>>> > > >>>> > > >>>> > On Mon, May 2, 2016 at 12:50 PM, Yong Zhang > >>>> > wrote: > >>>> >> > >>>> >> We are still not sure what is the problem, if you cannot show us > with > >>>> >> some > >>>> >> example data. > >>>> >> > >>>> >> For dps with 42632 rows, and swig with 42034 rows, if dps full > outer > >>>> >> join > >>>> >> with swig on 3 columns; with additional filters, get the same > >>>> >> resultSet row > >>>> >> count as dps lefter outer join with swig on 3 columns, with > >>>> >> additional > >>>> >> filters, again get the the same resultSet row count as dps right > >>>> >> outer join > >>>> >> with swig on 3 columns, with same additional filters. > >>>> >> > >>>> >> Without knowing your data, I cannot see the reason that has to be a > >>>> >> bug in > >>>> >> the spark. > >>>> >> > >>>> >> Am I misunderstanding your bug? > >>>> >> > >>>> >> Yong > >>>> >> > >>>> >> > >>>> >> From: kpe...@gmail.com > >>>> >> Date: Mon, 2 May 2016 12:11:18 -0700 > >>>> >> Subject: Re: Weird results with Spark SQL Outer joins > >>>> >> To: gourav.sengu...@gmail.com > >>>> >> CC: user@spark.apache.org > >>>> >> > >>>> >> > >>>> >> Gourav, > >>>> >> > >>>> >> I wish that was case, but I have done a select count on each of the > >>>> >> two > >>>> >> tables individually and they return back different number of rows: > >>>> >> > >>>> >> > >>>> >> dps.registerTempTable("dps_pin_promo_lt") > >>>> >> swig.registerTempTable("swig_pin_promo_lt") > >>>> >> > >>>> >> > >>>> >> dps.count() > >>>> >> RESULT: 42632 > >>>> >> > >>>> >> > >>>> >> swig.count() > >>>> >> RESULT: 42034 > >>>> >> > >>>> >> On Mon, May 2, 2016 at 11:55 AM, Gourav Sengupta > >>>> >> wrote: > >>>> >> > >>>> >> This shows that both the tables have matching records and no > >>>> >> mismatches. > >>>> >> Therefore obviously you have the same results irrespective of > whether > >>>> >> you > >>>> >> use right or left join. > >>>> >> > >>>> >> I think that there is no problem here, unless I am missing > something. > >>>> >> > >>>> >> Regards, > >>>> >> Gourav > >>>> >> > >>>> >> On Mon, May 2, 2016 at 7:48 PM, kpeng1 wrote: > >>>> >> > >>>> >> Also, the results of the inner query produced the same results: > >>>> >> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , > >>>> >> d.account > >>>> >> AS > >>>> >> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , > >>>> >> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN > >>>> >> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account > AND > >>>> >> s.ad > >>>> >> = > >>>> >> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= > >>>> >> '2016-01-03'").count() > >>>> >> RESULT:23747 > >>>> >> > >>>> >> > >>>> >> > >>>> >> -- > >>>> >> View this message in context: > >>>> >> > >>>> >> > http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html > >>>> >> Sent from the Apache Spark User List mailing list archive at > >>>> >> Nabble.com. > >>>> >> > >>>> >> > - > >>>> >> To unsubscribe, e-mail: user-unsubscr...@spark.apache.org > >>>> >> For additional commands, e-mail: user-h...@spark.apache.org > >>>> >> > >>>> >> > >>>> >> > >>>> > > >>> > >>> > >> > > > > > > > > -- > > Cesar Flores >
Re: Weird results with Spark SQL Outer joins
he outer join producing >>> less rows than the smallest table; if there are no matches it should still >>> bring in that row as part of the outer join. For the full and right outer >>> join I am expecting to see a minimum of res6 rows, but I get less, is there >>> something specific that I am missing here? I am expecting that the full >>> outer join would give me the union of the two table sets so I am expecting >>> at least 42533 rows not 23809. >>> >>> >>> Gourav, >>> >>> I just ran this result set on a new session with slightly newer data... >>> still seeing those results. >>> >>> >>> >>> Thanks, >>> >>> KP >>> >>> >>> On Mon, May 2, 2016 at 11:16 PM, Davies Liu >>> wrote: >>>> >>>> as @Gourav said, all the join with different join type show the same >>>> results, >>>> which meant that all the rows from left could match at least one row >>>> from right, >>>> all the rows from right could match at least one row from left, even >>>> the number of row from left does not equal that of right. >>>> >>>> This is correct result. >>>> >>>> On Mon, May 2, 2016 at 2:13 PM, Kevin Peng wrote: >>>> > Yong, >>>> > >>>> > Sorry, let explain my deduction; it is going be difficult to get a >>>> > sample >>>> > data out since the dataset I am using is proprietary. >>>> > >>>> > From the above set queries (ones mentioned in above comments) both >>>> > inner and >>>> > outer join are producing the same counts. They are basically pulling >>>> > out >>>> > selected columns from the query, but there is no roll up happening or >>>> > anything that would possible make it suspicious that there is any >>>> > difference >>>> > besides the type of joins. The tables are matched based on three keys >>>> > that >>>> > are present in both tables (ad, account, and date), on top of this >>>> > they are >>>> > filtered by date being above 2016-01-03. Since all the joins are >>>> > producing >>>> > the same counts, the natural suspicions is that the tables are >>>> > identical, >>>> > but I when I run the following two queries: >>>> > >>>> > scala> sqlContext.sql("select * from swig_pin_promo_lt where date >>>> >>='2016-01-03'").count >>>> > >>>> > res14: Long = 34158 >>>> > >>>> > scala> sqlContext.sql("select * from dps_pin_promo_lt where date >>>> >>='2016-01-03'").count >>>> > >>>> > res15: Long = 42693 >>>> > >>>> > >>>> > The above two queries filter out the data based on date used by the >>>> > joins of >>>> > 2016-01-03 and you can see the row count between the two tables are >>>> > different, which is why I am suspecting something is wrong with the >>>> > outer >>>> > joins in spark sql, because in this situation the right and outer >>>> > joins may >>>> > produce the same results, but it should not be equal to the left join >>>> > and >>>> > definitely not the inner join; unless I am missing something. >>>> > >>>> > >>>> > Side note: In my haste response above I posted the wrong counts for >>>> > dps.count, the real value is res16: Long = 42694 >>>> > >>>> > >>>> > Thanks, >>>> > >>>> > >>>> > KP >>>> > >>>> > >>>> > >>>> > >>>> > On Mon, May 2, 2016 at 12:50 PM, Yong Zhang >>>> > wrote: >>>> >> >>>> >> We are still not sure what is the problem, if you cannot show us with >>>> >> some >>>> >> example data. >>>> >> >>>> >> For dps with 42632 rows, and swig with 42034 rows, if dps full outer >>>> >> join >>>> >> with swig on 3 columns; with additional filters, get the same >>>> >> resultSet row >>>> >> count as dps lefter outer join with swig on 3 columns,
Re: Weird results with Spark SQL Outer joins
nks, >> >> KP >> >> >> On Mon, May 2, 2016 at 11:16 PM, Davies Liu >> wrote: >> >>> as @Gourav said, all the join with different join type show the same >>> results, >>> which meant that all the rows from left could match at least one row >>> from right, >>> all the rows from right could match at least one row from left, even >>> the number of row from left does not equal that of right. >>> >>> This is correct result. >>> >>> On Mon, May 2, 2016 at 2:13 PM, Kevin Peng wrote: >>> > Yong, >>> > >>> > Sorry, let explain my deduction; it is going be difficult to get a >>> sample >>> > data out since the dataset I am using is proprietary. >>> > >>> > From the above set queries (ones mentioned in above comments) both >>> inner and >>> > outer join are producing the same counts. They are basically pulling >>> out >>> > selected columns from the query, but there is no roll up happening or >>> > anything that would possible make it suspicious that there is any >>> difference >>> > besides the type of joins. The tables are matched based on three keys >>> that >>> > are present in both tables (ad, account, and date), on top of this >>> they are >>> > filtered by date being above 2016-01-03. Since all the joins are >>> producing >>> > the same counts, the natural suspicions is that the tables are >>> identical, >>> > but I when I run the following two queries: >>> > >>> > scala> sqlContext.sql("select * from swig_pin_promo_lt where date >>> >>='2016-01-03'").count >>> > >>> > res14: Long = 34158 >>> > >>> > scala> sqlContext.sql("select * from dps_pin_promo_lt where date >>> >>='2016-01-03'").count >>> > >>> > res15: Long = 42693 >>> > >>> > >>> > The above two queries filter out the data based on date used by the >>> joins of >>> > 2016-01-03 and you can see the row count between the two tables are >>> > different, which is why I am suspecting something is wrong with the >>> outer >>> > joins in spark sql, because in this situation the right and outer >>> joins may >>> > produce the same results, but it should not be equal to the left join >>> and >>> > definitely not the inner join; unless I am missing something. >>> > >>> > >>> > Side note: In my haste response above I posted the wrong counts for >>> > dps.count, the real value is res16: Long = 42694 >>> > >>> > >>> > Thanks, >>> > >>> > >>> > KP >>> > >>> > >>> > >>> > >>> > On Mon, May 2, 2016 at 12:50 PM, Yong Zhang >>> wrote: >>> >> >>> >> We are still not sure what is the problem, if you cannot show us with >>> some >>> >> example data. >>> >> >>> >> For dps with 42632 rows, and swig with 42034 rows, if dps full outer >>> join >>> >> with swig on 3 columns; with additional filters, get the same >>> resultSet row >>> >> count as dps lefter outer join with swig on 3 columns, with additional >>> >> filters, again get the the same resultSet row count as dps right >>> outer join >>> >> with swig on 3 columns, with same additional filters. >>> >> >>> >> Without knowing your data, I cannot see the reason that has to be a >>> bug in >>> >> the spark. >>> >> >>> >> Am I misunderstanding your bug? >>> >> >>> >> Yong >>> >> >>> >> >>> >> From: kpe...@gmail.com >>> >> Date: Mon, 2 May 2016 12:11:18 -0700 >>> >> Subject: Re: Weird results with Spark SQL Outer joins >>> >> To: gourav.sengu...@gmail.com >>> >> CC: user@spark.apache.org >>> >> >>> >> >>> >> Gourav, >>> >> >>> >> I wish that was case, but I have done a select count on each of the >>> two >>> >> tables individually and they return back different number of rows: >>> >> >>> >> >>> >> dps.registerTempTable("dps_pin_promo_lt") >>> >> swig.registerTempTable("swig_pin_promo_lt") >>> >> >>> >> >>> >> dps.count() >>> >> RESULT: 42632 >>> >> >>> >> >>> >> swig.count() >>> >> RESULT: 42034 >>> >> >>> >> On Mon, May 2, 2016 at 11:55 AM, Gourav Sengupta >>> >> wrote: >>> >> >>> >> This shows that both the tables have matching records and no >>> mismatches. >>> >> Therefore obviously you have the same results irrespective of whether >>> you >>> >> use right or left join. >>> >> >>> >> I think that there is no problem here, unless I am missing something. >>> >> >>> >> Regards, >>> >> Gourav >>> >> >>> >> On Mon, May 2, 2016 at 7:48 PM, kpeng1 wrote: >>> >> >>> >> Also, the results of the inner query produced the same results: >>> >> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , >>> d.account >>> >> AS >>> >> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >>> >> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN >>> >> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND >>> s.ad >>> >> = >>> >> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= >>> '2016-01-03'").count() >>> >> RESULT:23747 >>> >> >>> >> >>> >> >>> >> -- >>> >> View this message in context: >>> >> >>> http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html >>> >> Sent from the Apache Spark User List mailing list archive at >>> Nabble.com. >>> >> >>> >> - >>> >> To unsubscribe, e-mail: user-unsubscr...@spark.apache.org >>> >> For additional commands, e-mail: user-h...@spark.apache.org >>> >> >>> >> >>> >> >>> > >>> >> >> > -- Cesar Flores
Re: Weird results with Spark SQL Outer joins
I am using is proprietary. >> > >> > From the above set queries (ones mentioned in above comments) both >> inner and >> > outer join are producing the same counts. They are basically pulling >> out >> > selected columns from the query, but there is no roll up happening or >> > anything that would possible make it suspicious that there is any >> difference >> > besides the type of joins. The tables are matched based on three keys >> that >> > are present in both tables (ad, account, and date), on top of this they >> are >> > filtered by date being above 2016-01-03. Since all the joins are >> producing >> > the same counts, the natural suspicions is that the tables are >> identical, >> > but I when I run the following two queries: >> > >> > scala> sqlContext.sql("select * from swig_pin_promo_lt where date >> >>='2016-01-03'").count >> > >> > res14: Long = 34158 >> > >> > scala> sqlContext.sql("select * from dps_pin_promo_lt where date >> >>='2016-01-03'").count >> > >> > res15: Long = 42693 >> > >> > >> > The above two queries filter out the data based on date used by the >> joins of >> > 2016-01-03 and you can see the row count between the two tables are >> > different, which is why I am suspecting something is wrong with the >> outer >> > joins in spark sql, because in this situation the right and outer joins >> may >> > produce the same results, but it should not be equal to the left join >> and >> > definitely not the inner join; unless I am missing something. >> > >> > >> > Side note: In my haste response above I posted the wrong counts for >> > dps.count, the real value is res16: Long = 42694 >> > >> > >> > Thanks, >> > >> > >> > KP >> > >> > >> > >> > >> > On Mon, May 2, 2016 at 12:50 PM, Yong Zhang >> wrote: >> >> >> >> We are still not sure what is the problem, if you cannot show us with >> some >> >> example data. >> >> >> >> For dps with 42632 rows, and swig with 42034 rows, if dps full outer >> join >> >> with swig on 3 columns; with additional filters, get the same >> resultSet row >> >> count as dps lefter outer join with swig on 3 columns, with additional >> >> filters, again get the the same resultSet row count as dps right outer >> join >> >> with swig on 3 columns, with same additional filters. >> >> >> >> Without knowing your data, I cannot see the reason that has to be a >> bug in >> >> the spark. >> >> >> >> Am I misunderstanding your bug? >> >> >> >> Yong >> >> >> >> >> >> From: kpe...@gmail.com >> >> Date: Mon, 2 May 2016 12:11:18 -0700 >> >> Subject: Re: Weird results with Spark SQL Outer joins >> >> To: gourav.sengu...@gmail.com >> >> CC: user@spark.apache.org >> >> >> >> >> >> Gourav, >> >> >> >> I wish that was case, but I have done a select count on each of the two >> >> tables individually and they return back different number of rows: >> >> >> >> >> >> dps.registerTempTable("dps_pin_promo_lt") >> >> swig.registerTempTable("swig_pin_promo_lt") >> >> >> >> >> >> dps.count() >> >> RESULT: 42632 >> >> >> >> >> >> swig.count() >> >> RESULT: 42034 >> >> >> >> On Mon, May 2, 2016 at 11:55 AM, Gourav Sengupta >> >> wrote: >> >> >> >> This shows that both the tables have matching records and no >> mismatches. >> >> Therefore obviously you have the same results irrespective of whether >> you >> >> use right or left join. >> >> >> >> I think that there is no problem here, unless I am missing something. >> >> >> >> Regards, >> >> Gourav >> >> >> >> On Mon, May 2, 2016 at 7:48 PM, kpeng1 wrote: >> >> >> >> Also, the results of the inner query produced the same results: >> >> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , >> d.account >> >> AS >> >> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >> >> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN >> >> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND >> s.ad >> >> = >> >> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= >> '2016-01-03'").count() >> >> RESULT:23747 >> >> >> >> >> >> >> >> -- >> >> View this message in context: >> >> >> http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html >> >> Sent from the Apache Spark User List mailing list archive at >> Nabble.com. >> >> >> >> - >> >> To unsubscribe, e-mail: user-unsubscr...@spark.apache.org >> >> For additional commands, e-mail: user-h...@spark.apache.org >> >> >> >> >> >> >> > >> > >
Re: Weird results with Spark SQL Outer joins
t; > scala> sqlContext.sql("select * from swig_pin_promo_lt where date > >>='2016-01-03'").count > > > > res14: Long = 34158 > > > > scala> sqlContext.sql("select * from dps_pin_promo_lt where date > >>='2016-01-03'").count > > > > res15: Long = 42693 > > > > > > The above two queries filter out the data based on date used by the > joins of > > 2016-01-03 and you can see the row count between the two tables are > > different, which is why I am suspecting something is wrong with the outer > > joins in spark sql, because in this situation the right and outer joins > may > > produce the same results, but it should not be equal to the left join and > > definitely not the inner join; unless I am missing something. > > > > > > Side note: In my haste response above I posted the wrong counts for > > dps.count, the real value is res16: Long = 42694 > > > > > > Thanks, > > > > > > KP > > > > > > > > > > On Mon, May 2, 2016 at 12:50 PM, Yong Zhang > wrote: > >> > >> We are still not sure what is the problem, if you cannot show us with > some > >> example data. > >> > >> For dps with 42632 rows, and swig with 42034 rows, if dps full outer > join > >> with swig on 3 columns; with additional filters, get the same resultSet > row > >> count as dps lefter outer join with swig on 3 columns, with additional > >> filters, again get the the same resultSet row count as dps right outer > join > >> with swig on 3 columns, with same additional filters. > >> > >> Without knowing your data, I cannot see the reason that has to be a bug > in > >> the spark. > >> > >> Am I misunderstanding your bug? > >> > >> Yong > >> > >> > >> From: kpe...@gmail.com > >> Date: Mon, 2 May 2016 12:11:18 -0700 > >> Subject: Re: Weird results with Spark SQL Outer joins > >> To: gourav.sengu...@gmail.com > >> CC: user@spark.apache.org > >> > >> > >> Gourav, > >> > >> I wish that was case, but I have done a select count on each of the two > >> tables individually and they return back different number of rows: > >> > >> > >> dps.registerTempTable("dps_pin_promo_lt") > >> swig.registerTempTable("swig_pin_promo_lt") > >> > >> > >> dps.count() > >> RESULT: 42632 > >> > >> > >> swig.count() > >> RESULT: 42034 > >> > >> On Mon, May 2, 2016 at 11:55 AM, Gourav Sengupta > >> wrote: > >> > >> This shows that both the tables have matching records and no mismatches. > >> Therefore obviously you have the same results irrespective of whether > you > >> use right or left join. > >> > >> I think that there is no problem here, unless I am missing something. > >> > >> Regards, > >> Gourav > >> > >> On Mon, May 2, 2016 at 7:48 PM, kpeng1 wrote: > >> > >> Also, the results of the inner query produced the same results: > >> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , > d.account > >> AS > >> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , > >> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN > >> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND > s.ad > >> = > >> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= > '2016-01-03'").count() > >> RESULT:23747 > >> > >> > >> > >> -- > >> View this message in context: > >> > http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html > >> Sent from the Apache Spark User List mailing list archive at Nabble.com. > >> > >> - > >> To unsubscribe, e-mail: user-unsubscr...@spark.apache.org > >> For additional commands, e-mail: user-h...@spark.apache.org > >> > >> > >> > > >
Re: Weird results with Spark SQL Outer joins
as @Gourav said, all the join with different join type show the same results, which meant that all the rows from left could match at least one row from right, all the rows from right could match at least one row from left, even the number of row from left does not equal that of right. This is correct result. On Mon, May 2, 2016 at 2:13 PM, Kevin Peng wrote: > Yong, > > Sorry, let explain my deduction; it is going be difficult to get a sample > data out since the dataset I am using is proprietary. > > From the above set queries (ones mentioned in above comments) both inner and > outer join are producing the same counts. They are basically pulling out > selected columns from the query, but there is no roll up happening or > anything that would possible make it suspicious that there is any difference > besides the type of joins. The tables are matched based on three keys that > are present in both tables (ad, account, and date), on top of this they are > filtered by date being above 2016-01-03. Since all the joins are producing > the same counts, the natural suspicions is that the tables are identical, > but I when I run the following two queries: > > scala> sqlContext.sql("select * from swig_pin_promo_lt where date >>='2016-01-03'").count > > res14: Long = 34158 > > scala> sqlContext.sql("select * from dps_pin_promo_lt where date >>='2016-01-03'").count > > res15: Long = 42693 > > > The above two queries filter out the data based on date used by the joins of > 2016-01-03 and you can see the row count between the two tables are > different, which is why I am suspecting something is wrong with the outer > joins in spark sql, because in this situation the right and outer joins may > produce the same results, but it should not be equal to the left join and > definitely not the inner join; unless I am missing something. > > > Side note: In my haste response above I posted the wrong counts for > dps.count, the real value is res16: Long = 42694 > > > Thanks, > > > KP > > > > > On Mon, May 2, 2016 at 12:50 PM, Yong Zhang wrote: >> >> We are still not sure what is the problem, if you cannot show us with some >> example data. >> >> For dps with 42632 rows, and swig with 42034 rows, if dps full outer join >> with swig on 3 columns; with additional filters, get the same resultSet row >> count as dps lefter outer join with swig on 3 columns, with additional >> filters, again get the the same resultSet row count as dps right outer join >> with swig on 3 columns, with same additional filters. >> >> Without knowing your data, I cannot see the reason that has to be a bug in >> the spark. >> >> Am I misunderstanding your bug? >> >> Yong >> >> >> From: kpe...@gmail.com >> Date: Mon, 2 May 2016 12:11:18 -0700 >> Subject: Re: Weird results with Spark SQL Outer joins >> To: gourav.sengu...@gmail.com >> CC: user@spark.apache.org >> >> >> Gourav, >> >> I wish that was case, but I have done a select count on each of the two >> tables individually and they return back different number of rows: >> >> >> dps.registerTempTable("dps_pin_promo_lt") >> swig.registerTempTable("swig_pin_promo_lt") >> >> >> dps.count() >> RESULT: 42632 >> >> >> swig.count() >> RESULT: 42034 >> >> On Mon, May 2, 2016 at 11:55 AM, Gourav Sengupta >> wrote: >> >> This shows that both the tables have matching records and no mismatches. >> Therefore obviously you have the same results irrespective of whether you >> use right or left join. >> >> I think that there is no problem here, unless I am missing something. >> >> Regards, >> Gourav >> >> On Mon, May 2, 2016 at 7:48 PM, kpeng1 wrote: >> >> Also, the results of the inner query produced the same results: >> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account >> AS >> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN >> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND s.ad >> = >> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= '2016-01-03'").count() >> RESULT:23747 >> >> >> >> -- >> View this message in context: >> http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html >> Sent from the Apache Spark User List mailing list archive at Nabble.com. >> >> - >> To unsubscribe, e-mail: user-unsubscr...@spark.apache.org >> For additional commands, e-mail: user-h...@spark.apache.org >> >> >> > - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Weird results with Spark SQL Outer joins
Hi, its morning 4:40 here, therefore I might not be getting things right. But there is a very high chance of getting spurious results in case you have created that variable more than once in IPython or pyspark shell and cached it and are re using it. Please close the sessions and create the variable only once and then add the table (after ensuring that you have dropped if first) and then check it. I had faced this issue once and then realized it was because of the above reasons. Regards, Gourav On Mon, May 2, 2016 at 10:13 PM, Kevin Peng wrote: > Yong, > > Sorry, let explain my deduction; it is going be difficult to get a sample > data out since the dataset I am using is proprietary. > > From the above set queries (ones mentioned in above comments) both inner > and outer join are producing the same counts. They are basically pulling > out selected columns from the query, but there is no roll up happening or > anything that would possible make it suspicious that there is any > difference besides the type of joins. The tables are matched based on > three keys that are present in both tables (ad, account, and date), on top > of this they are filtered by date being above 2016-01-03. Since all the > joins are producing the same counts, the natural suspicions is that the > tables are identical, but I when I run the following two queries: > > scala> sqlContext.sql("select * from swig_pin_promo_lt where date > >='2016-01-03'").count > > res14: Long = 34158 > > scala> sqlContext.sql("select * from dps_pin_promo_lt where date > >='2016-01-03'").count > > res15: Long = 42693 > > > The above two queries filter out the data based on date used by the joins > of 2016-01-03 and you can see the row count between the two tables are > different, which is why I am suspecting something is wrong with the outer > joins in spark sql, because in this situation the right and outer joins may > produce the same results, but it should not be equal to the left join and > definitely not the inner join; unless I am missing something. > > > Side note: In my haste response above I posted the wrong counts for > dps.count, the real value is res16: Long = 42694 > > > Thanks, > > > KP > > > > On Mon, May 2, 2016 at 12:50 PM, Yong Zhang wrote: > >> We are still not sure what is the problem, if you cannot show us with >> some example data. >> >> For dps with 42632 rows, and swig with 42034 rows, if dps full outer join >> with swig on 3 columns; with additional filters, get the same resultSet row >> count as dps lefter outer join with swig on 3 columns, with additional >> filters, again get the the same resultSet row count as dps right outer join >> with swig on 3 columns, with same additional filters. >> >> Without knowing your data, I cannot see the reason that has to be a bug >> in the spark. >> >> Am I misunderstanding your bug? >> >> Yong >> >> -- >> From: kpe...@gmail.com >> Date: Mon, 2 May 2016 12:11:18 -0700 >> Subject: Re: Weird results with Spark SQL Outer joins >> To: gourav.sengu...@gmail.com >> CC: user@spark.apache.org >> >> >> Gourav, >> >> I wish that was case, but I have done a select count on each of the two >> tables individually and they return back different number of rows: >> >> >> dps.registerTempTable("dps_pin_promo_lt") >> swig.registerTempTable("swig_pin_promo_lt") >> >> >> dps.count() >> RESULT: 42632 >> >> >> swig.count() >> RESULT: 42034 >> >> On Mon, May 2, 2016 at 11:55 AM, Gourav Sengupta < >> gourav.sengu...@gmail.com> wrote: >> >> This shows that both the tables have matching records and no mismatches. >> Therefore obviously you have the same results irrespective of whether you >> use right or left join. >> >> I think that there is no problem here, unless I am missing something. >> >> Regards, >> Gourav >> >> On Mon, May 2, 2016 at 7:48 PM, kpeng1 wrote: >> >> Also, the results of the inner query produced the same results: >> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account >> AS >> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN >> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND >> s.ad = >> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= >> '2016-01-03'").count() >> RESULT:23747 >> >> >> >> -- >> View this message in context: >> http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html >> Sent from the Apache Spark User List mailing list archive at Nabble.com. >> >> - >> To unsubscribe, e-mail: user-unsubscr...@spark.apache.org >> For additional commands, e-mail: user-h...@spark.apache.org >> >> >> >> >
Re: Weird results with Spark SQL Outer joins
Yong, Sorry, let explain my deduction; it is going be difficult to get a sample data out since the dataset I am using is proprietary. >From the above set queries (ones mentioned in above comments) both inner and outer join are producing the same counts. They are basically pulling out selected columns from the query, but there is no roll up happening or anything that would possible make it suspicious that there is any difference besides the type of joins. The tables are matched based on three keys that are present in both tables (ad, account, and date), on top of this they are filtered by date being above 2016-01-03. Since all the joins are producing the same counts, the natural suspicions is that the tables are identical, but I when I run the following two queries: scala> sqlContext.sql("select * from swig_pin_promo_lt where date >='2016-01-03'").count res14: Long = 34158 scala> sqlContext.sql("select * from dps_pin_promo_lt where date >='2016-01-03'").count res15: Long = 42693 The above two queries filter out the data based on date used by the joins of 2016-01-03 and you can see the row count between the two tables are different, which is why I am suspecting something is wrong with the outer joins in spark sql, because in this situation the right and outer joins may produce the same results, but it should not be equal to the left join and definitely not the inner join; unless I am missing something. Side note: In my haste response above I posted the wrong counts for dps.count, the real value is res16: Long = 42694 Thanks, KP On Mon, May 2, 2016 at 12:50 PM, Yong Zhang wrote: > We are still not sure what is the problem, if you cannot show us with some > example data. > > For dps with 42632 rows, and swig with 42034 rows, if dps full outer join > with swig on 3 columns; with additional filters, get the same resultSet row > count as dps lefter outer join with swig on 3 columns, with additional > filters, again get the the same resultSet row count as dps right outer join > with swig on 3 columns, with same additional filters. > > Without knowing your data, I cannot see the reason that has to be a bug in > the spark. > > Am I misunderstanding your bug? > > Yong > > -------------- > From: kpe...@gmail.com > Date: Mon, 2 May 2016 12:11:18 -0700 > Subject: Re: Weird results with Spark SQL Outer joins > To: gourav.sengu...@gmail.com > CC: user@spark.apache.org > > > Gourav, > > I wish that was case, but I have done a select count on each of the two > tables individually and they return back different number of rows: > > > dps.registerTempTable("dps_pin_promo_lt") > swig.registerTempTable("swig_pin_promo_lt") > > > dps.count() > RESULT: 42632 > > > swig.count() > RESULT: 42034 > > On Mon, May 2, 2016 at 11:55 AM, Gourav Sengupta < > gourav.sengu...@gmail.com> wrote: > > This shows that both the tables have matching records and no mismatches. > Therefore obviously you have the same results irrespective of whether you > use right or left join. > > I think that there is no problem here, unless I am missing something. > > Regards, > Gourav > > On Mon, May 2, 2016 at 7:48 PM, kpeng1 wrote: > > Also, the results of the inner query produced the same results: > sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account > AS > d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , > d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN > dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND s.ad > = > d.ad) WHERE s.date >= '2016-01-03'AND d.date >= '2016-01-03'").count() > RESULT:23747 > > > > -- > View this message in context: > http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html > Sent from the Apache Spark User List mailing list archive at Nabble.com. > > - > To unsubscribe, e-mail: user-unsubscr...@spark.apache.org > For additional commands, e-mail: user-h...@spark.apache.org > > > >
RE: Weird results with Spark SQL Outer joins
We are still not sure what is the problem, if you cannot show us with some example data. For dps with 42632 rows, and swig with 42034 rows, if dps full outer join with swig on 3 columns; with additional filters, get the same resultSet row count as dps lefter outer join with swig on 3 columns, with additional filters, again get the the same resultSet row count as dps right outer join with swig on 3 columns, with same additional filters. Without knowing your data, I cannot see the reason that has to be a bug in the spark. Am I misunderstanding your bug? Yong From: kpe...@gmail.com Date: Mon, 2 May 2016 12:11:18 -0700 Subject: Re: Weird results with Spark SQL Outer joins To: gourav.sengu...@gmail.com CC: user@spark.apache.org Gourav, I wish that was case, but I have done a select count on each of the two tables individually and they return back different number of rows: dps.registerTempTable("dps_pin_promo_lt") swig.registerTempTable("swig_pin_promo_lt") dps.count() RESULT: 42632 swig.count() RESULT: 42034 On Mon, May 2, 2016 at 11:55 AM, Gourav Sengupta wrote: This shows that both the tables have matching records and no mismatches. Therefore obviously you have the same results irrespective of whether you use right or left join. I think that there is no problem here, unless I am missing something. Regards,Gourav On Mon, May 2, 2016 at 7:48 PM, kpeng1 wrote: Also, the results of the inner query produced the same results: sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account AS d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND s.ad = d.ad) WHERE s.date >= '2016-01-03'AND d.date >= '2016-01-03'").count() RESULT:23747 -- View this message in context: http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html Sent from the Apache Spark User List mailing list archive at Nabble.com. - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Weird results with Spark SQL Outer joins
Gourav, I wish that was case, but I have done a select count on each of the two tables individually and they return back different number of rows: dps.registerTempTable("dps_pin_promo_lt") swig.registerTempTable("swig_pin_promo_lt") dps.count() RESULT: 42632 swig.count() RESULT: 42034 On Mon, May 2, 2016 at 11:55 AM, Gourav Sengupta wrote: > This shows that both the tables have matching records and no mismatches. > Therefore obviously you have the same results irrespective of whether you > use right or left join. > > I think that there is no problem here, unless I am missing something. > > Regards, > Gourav > > On Mon, May 2, 2016 at 7:48 PM, kpeng1 wrote: > >> Also, the results of the inner query produced the same results: >> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account >> AS >> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN >> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND >> s.ad = >> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= >> '2016-01-03'").count() >> RESULT:23747 >> >> >> >> -- >> View this message in context: >> http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html >> Sent from the Apache Spark User List mailing list archive at Nabble.com. >> >> - >> To unsubscribe, e-mail: user-unsubscr...@spark.apache.org >> For additional commands, e-mail: user-h...@spark.apache.org >> >> >
Re: Weird results with Spark SQL Outer joins
This shows that both the tables have matching records and no mismatches. Therefore obviously you have the same results irrespective of whether you use right or left join. I think that there is no problem here, unless I am missing something. Regards, Gourav On Mon, May 2, 2016 at 7:48 PM, kpeng1 wrote: > Also, the results of the inner query produced the same results: > sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account > AS > d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , > d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN > dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND s.ad > = > d.ad) WHERE s.date >= '2016-01-03'AND d.date >= '2016-01-03'").count() > RESULT:23747 > > > > -- > View this message in context: > http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html > Sent from the Apache Spark User List mailing list archive at Nabble.com. > > - > To unsubscribe, e-mail: user-unsubscr...@spark.apache.org > For additional commands, e-mail: user-h...@spark.apache.org > >
Re: Weird results with Spark SQL Outer joins
Also, the results of the inner query produced the same results: sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account AS d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s INNER JOIN dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND s.ad = d.ad) WHERE s.date >= '2016-01-03'AND d.date >= '2016-01-03'").count() RESULT:23747 -- View this message in context: http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861p26863.html Sent from the Apache Spark User List mailing list archive at Nabble.com. - To unsubscribe, e-mail: user-unsubscr...@spark.apache.org For additional commands, e-mail: user-h...@spark.apache.org
Re: Weird results with Spark SQL Outer joins
Hi Kevin, Thanks. Please post the result of the same query with INNER JOIN and then it will give us a bit of insight. Regards, Gourav On Mon, May 2, 2016 at 7:10 PM, Kevin Peng wrote: > Gourav, > > Apologies. I edited my post with this information: > Spark version: 1.6 > Result from spark shell > OS: Linux version 2.6.32-431.20.3.el6.x86_64 ( > mockbu...@c6b9.bsys.dev.centos.org) (gcc version 4.4.7 20120313 (Red Hat > 4.4.7-4) (GCC) ) #1 SMP Thu Jun 19 21:14:45 UTC 2014 > > Thanks, > > KP > > On Mon, May 2, 2016 at 11:05 AM, Gourav Sengupta < > gourav.sengu...@gmail.com> wrote: > >> Hi, >> >> As always, can you please write down details regarding your SPARK cluster >> - the version, OS, IDE used, etc? >> >> Regards, >> Gourav Sengupta >> >> On Mon, May 2, 2016 at 5:58 PM, kpeng1 wrote: >> >>> Hi All, >>> >>> I am running into a weird result with Spark SQL Outer joins. The results >>> for all of them seem to be the same, which does not make sense due to the >>> data. Here are the queries that I am running with the results: >>> >>> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , >>> d.account AS >>> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >>> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s FULL OUTER JOIN >>> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND >>> s.ad = >>> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= >>> '2016-01-03'").count() >>> RESULT:23747 >>> >>> >>> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , >>> d.account AS >>> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >>> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s LEFT OUTER JOIN >>> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND >>> s.ad = >>> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= >>> '2016-01-03'").count() >>> RESULT:23747 >>> >>> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , >>> d.account AS >>> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >>> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s RIGHT OUTER JOIN >>> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND >>> s.ad = >>> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= >>> '2016-01-03'").count() >>> RESULT: 23747 >>> >>> Was wondering if someone had encountered this issues before. >>> >>> >>> >>> -- >>> View this message in context: >>> http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861.html >>> Sent from the Apache Spark User List mailing list archive at Nabble.com. >>> >>> - >>> To unsubscribe, e-mail: user-unsubscr...@spark.apache.org >>> For additional commands, e-mail: user-h...@spark.apache.org >>> >>> >> >
Re: Weird results with Spark SQL Outer joins
Gourav, Apologies. I edited my post with this information: Spark version: 1.6 Result from spark shell OS: Linux version 2.6.32-431.20.3.el6.x86_64 ( mockbu...@c6b9.bsys.dev.centos.org) (gcc version 4.4.7 20120313 (Red Hat 4.4.7-4) (GCC) ) #1 SMP Thu Jun 19 21:14:45 UTC 2014 Thanks, KP On Mon, May 2, 2016 at 11:05 AM, Gourav Sengupta wrote: > Hi, > > As always, can you please write down details regarding your SPARK cluster > - the version, OS, IDE used, etc? > > Regards, > Gourav Sengupta > > On Mon, May 2, 2016 at 5:58 PM, kpeng1 wrote: > >> Hi All, >> >> I am running into a weird result with Spark SQL Outer joins. The results >> for all of them seem to be the same, which does not make sense due to the >> data. Here are the queries that I am running with the results: >> >> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account >> AS >> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s FULL OUTER JOIN >> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND >> s.ad = >> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= >> '2016-01-03'").count() >> RESULT:23747 >> >> >> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account >> AS >> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s LEFT OUTER JOIN >> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND >> s.ad = >> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= >> '2016-01-03'").count() >> RESULT:23747 >> >> sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account >> AS >> d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , >> d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s RIGHT OUTER JOIN >> dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND >> s.ad = >> d.ad) WHERE s.date >= '2016-01-03'AND d.date >= >> '2016-01-03'").count() >> RESULT: 23747 >> >> Was wondering if someone had encountered this issues before. >> >> >> >> -- >> View this message in context: >> http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861.html >> Sent from the Apache Spark User List mailing list archive at Nabble.com. >> >> - >> To unsubscribe, e-mail: user-unsubscr...@spark.apache.org >> For additional commands, e-mail: user-h...@spark.apache.org >> >> >
Re: Weird results with Spark SQL Outer joins
Hi, As always, can you please write down details regarding your SPARK cluster - the version, OS, IDE used, etc? Regards, Gourav Sengupta On Mon, May 2, 2016 at 5:58 PM, kpeng1 wrote: > Hi All, > > I am running into a weird result with Spark SQL Outer joins. The results > for all of them seem to be the same, which does not make sense due to the > data. Here are the queries that I am running with the results: > > sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account > AS > d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , > d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s FULL OUTER JOIN > dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND s.ad > = > d.ad) WHERE s.date >= '2016-01-03'AND d.date >= '2016-01-03'").count() > RESULT:23747 > > > sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account > AS > d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , > d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s LEFT OUTER JOIN > dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND s.ad > = > d.ad) WHERE s.date >= '2016-01-03'AND d.date >= '2016-01-03'").count() > RESULT:23747 > > sqlContext.sql("SELECT s.date AS edate , s.account AS s_acc , d.account > AS > d_acc , s.ad as s_ad , d.ad as d_ad , s.spend AS s_spend , > d.spend_in_dollar AS d_spend FROM swig_pin_promo_lt s RIGHT OUTER JOIN > dps_pin_promo_lt d ON (s.date = d.date AND s.account = d.account AND s.ad > = > d.ad) WHERE s.date >= '2016-01-03'AND d.date >= '2016-01-03'").count() > RESULT: 23747 > > Was wondering if someone had encountered this issues before. > > > > -- > View this message in context: > http://apache-spark-user-list.1001560.n3.nabble.com/Weird-results-with-Spark-SQL-Outer-joins-tp26861.html > Sent from the Apache Spark User List mailing list archive at Nabble.com. > > - > To unsubscribe, e-mail: user-unsubscr...@spark.apache.org > For additional commands, e-mail: user-h...@spark.apache.org > >