Support for Windowing and Analytics functions in Spark SQL

2015-06-22 Thread Sourav Mazumder
Hi,

Though the documentation does not explicitly mention support for Windowing
and Analytics function in Spark SQL, looks like it is not supported.

I tried running a query like Select Lead(column name, 1) over (Partition
By column name order by column name) from table name and I got error
saying that this feature is unsupported.

I tried it in Databricks cloud and that supports Spark 1.4.

Can anyone please confirm this ?

Regards,
Sourav


Re: Support for Windowing and Analytics functions in Spark SQL

2015-06-22 Thread ayan guha
1.4 supports it
On 23 Jun 2015 02:59, Sourav Mazumder sourav.mazumde...@gmail.com wrote:

 Hi,

 Though the documentation does not explicitly mention support for Windowing
 and Analytics function in Spark SQL, looks like it is not supported.

 I tried running a query like Select Lead(column name, 1) over (Partition
 By column name order by column name) from table name and I got error
 saying that this feature is unsupported.

 I tried it in Databricks cloud and that supports Spark 1.4.

 Can anyone please confirm this ?

 Regards,
 Sourav



RE: Support for Windowing and Analytics functions in Spark SQL

2015-06-22 Thread Cheng, Hao
Yes, with should be with HiveContext, not SQLContext.

From: ayan guha [mailto:guha.a...@gmail.com]
Sent: Tuesday, June 23, 2015 2:51 AM
To: smazumder
Cc: user
Subject: Re: Support for Windowing and Analytics functions in Spark SQL


1.4 supports it
On 23 Jun 2015 02:59, Sourav Mazumder 
sourav.mazumde...@gmail.commailto:sourav.mazumde...@gmail.com wrote:
Hi,
Though the documentation does not explicitly mention support for Windowing and 
Analytics function in Spark SQL, looks like it is not supported.
I tried running a query like Select Lead(column name, 1) over (Partition By 
column name order by column name) from table name and I got error saying 
that this feature is unsupported.
I tried it in Databricks cloud and that supports Spark 1.4.
Can anyone please confirm this ?
Regards,
Sourav