Re: funneling/drill down

[Joanna He]

Ali,

I think the Intersect Count Measure (
http://kylin.apache.org/blog/2016/11/28/intersect-count/) can help you with
your funnel analysis, although I don't entirely understand your use case.
If you could give me an example of your table structure, I could help more.

On Sun, Apr 12, 2020 at 10:30 AM ar174  wrote:

> Hi,
>
> I’m doing the initial assessment of Kylin’s suitability for my use case,
> which is to interactively explore events that have happened to a group
> patients to show aggregates such as counts and total costs.
>
> I’ve setup Kyline with docker and I’ve created a cube and the aggregate
> calculations are performing very well. What I am struggling to do is
> tracking patients from one event to another. Please see the image below for
> an explanation.
>
> 
>
> Events 1, 2 & 3 are rows in the database and have a patient id to link
> them.
> Each row also have various fields that the user (in the UI) can use to
> filter with such as location, diagnosis, gender, etc. My use case is that
> if
> I filter patients in the first event by dimensions A and then want to
> filters those patient by dimensions D in the second event and so and then
> generate aggregates for each event. Is Kylin suitable for this(and if not,
> any alternatives?) and how can I do this?
>
> Thanks,
> Ali
>
>
> --
> Sent from: http://apache-kylin.74782.x6.nabble.com/
>

funneling/drill down

[ar174]

Hi,

I’m doing the initial assessment of Kylin’s suitability for my use case,
which is to interactively explore events that have happened to a group
patients to show aggregates such as counts and total costs.

I’ve setup Kyline with docker and I’ve created a cube and the aggregate
calculations are performing very well. What I am struggling to do is
tracking patients from one event to another. Please see the image below for
an explanation.

 

Events 1, 2 & 3 are rows in the database and have a patient id to link them.
Each row also have various fields that the user (in the UI) can use to
filter with such as location, diagnosis, gender, etc. My use case is that if
I filter patients in the first event by dimensions A and then want to
filters those patient by dimensions D in the second event and so and then
generate aggregates for each event. Is Kylin suitable for this(and if not,
any alternatives?) and how can I do this?

Thanks,
Ali


--
Sent from: http://apache-kylin.74782.x6.nabble.com/

Re:Kylin GUI error "Cannot get HiveTableMeta" on AWS EMR with Glue as hive metastore

[Xiaoxiang Yu]

Thank you for update on this. And appreciate Kaige 's contribution.










--

Best wishes to you ! 
From ：Xiaoxiang Yu





At 2020-04-10 22:38:00, "mvishnubhatta"  
wrote:
>Great. Thank you. I tried out Kaige's suggestion on
>https://issues.apache.org/jira/browse/KYLIN-3685 that you pointed me to and
>that worked for me.
>
>--
>Sent from: http://apache-kylin.74782.x6.nabble.com/