spark-user mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From Faiz Chachiya <faiz.in...@gmail.com>
Subject Re: Spark DataFrame/DataSet Wide Transformations
Date Thu, 07 Feb 2019 05:55:51 GMT
Hi Hemant - Well it is pretty clear to me that conceptually the
transformations would behave in similar way.

My question is how to identify the parent dependencies as you would
typically do with RDD.

Thanks,
Faiz

On Thu, Feb 7, 2019 at 10:22 AM hemant singh <hemant2184@gmail.com> wrote:

> Same concept applies to Dataframe as it is with RDD with respect to
> transformations. Both are distributed data set.
>
> Thanks
>
> On Thu, Feb 7, 2019 at 8:51 AM Faiz Chachiya <faiz.india@gmail.com> wrote:
>
>> Hello Team,
>>
>> With RDDs it is pretty clear which operations would result in wide
>> transformations and there are also options available to find out parent
>> dependencies
>>
>> I have been struggling to do the same with DataFrame/DataSet, I need your
>> helping in finding out which operations may lead to wide transformations
>> like (OrderBy) and if there is way to find out the parent dependencies.
>>
>> There is one way to find out parent dependencies by converting the DF/DS
>> to RDD and invoke the dependencies.
>>
>> I hope my question is clear and would request your help with it.
>>
>> Thanks,
>> Faiz
>>
>

Mime
View raw message