Hi William, Yes, we need to have a set of single measures for a dataset and then the average of this set of measures without having to recalculate each single measure, because we already have to value of the measure and we only need to calculate the average.
Regards -----Original Message----- From: William Guo <[email protected]> Sent: Thursday, September 26, 2019 4:29 PM To: [email protected] Subject: Re: Average of the measures hi, Based on your message, I understand to calculate average , you want to find a way to avoid recalculate the average measure when datasets are updated. Tell me whether my understanding is right or not? Thanks, William On Wed, Sep 25, 2019 at 11:17 PM < [email protected]> wrote: > Hi William, > > The use case is the following, we have a datalake that is structured > in datasets, each one of these datasets can have a set of quality > measures and user wants to have a global measure of the dataset > quality that is the average of all dataset quality measures. To do > this what we have done is defining a custom measure that calculates > this average, but it implies calculating again all quality measures > and we were trying to find a way of calculating the average without > recalculating quality measures. > > Regards > > -----Original Message----- > From: William Guo <[email protected]> > Sent: Tuesday, September 24, 2019 3:14 AM > To: [email protected] > Subject: Re: Average of the measures > > hi, > > Could you tell us your use case? > Normally, you can use avg function from spark sql. > Griffin support spark sql directly. > > Thanks, > William > > On Thu, Sep 19, 2019 at 6:50 PM < > [email protected]> wrote: > > > Hello, > > > > We need to create an average of the measures for a certain data set, > > has anybody done this with Apache Griffin? > > > > Regards > > >
