
Re: Set analysis
Sunny Talwar Jul 31, 2015 7:23 AM (in response to Satish Pawar)May be try this:
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)/
Sum(Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_type, branch_name))


Re: Set analysis
Sunny Talwar Jul 31, 2015 7:35 AM (in response to Satish Pawar)Makes sense. How are you drilling into a different dimension as there is no groups available in Qlik Sense (if I am not wrong)? Are you using variable to switch from one dimension to another?

Re: Set analysis
Satish Pawar Jul 31, 2015 7:42 AM (in response to Sunny Talwar )No i'm not using variable.
there is one option from where we can create group for drill down and also manipulate in Chart.
Qlik Sense  Drilldown Group Dimension (video)
but there is no option to add such group into expression.
its quite strange

Re: Set analysis
Sunny Talwar Jul 31, 2015 7:49 AM (in response to Satish Pawar)I guess I understand the issue, but I am new to Qlik Sense as well. May be somebody from Qlik Tech might be able to help you better.
Best,
Sunny

Re: Set analysis
Satish Pawar Jul 31, 2015 7:56 AM (in response to Sunny Talwar )here
i'm trying to get the row wise count at denominator.
if you have some new way.
please suggest.

Re: Set analysis
Ruben Marin Jul 31, 2015 8:53 AM (in response to Satish Pawar)Hi, just guessing, maybe:
If(Count(Distinct branch_name)>1,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)/
Sum(Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_type),
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)/
Sum(Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value), branch_name))
)
I'm trying to check where the dimension is branch_name or branch_type, and each has his aggr. If some branch_type only has 1 branch_name there should be no difference between using branch_type or branch_name for aggr().


Re: Set analysis
Ruben Marin Aug 3, 2015 6:52 AM (in response to Satish Pawar)Hi Satish, maybe this expression:
If(Count(Distinct branch_name)>1,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Sum(Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_type),
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)/
/Sum(count(TOTAL {<param_name={'TEMP'}, param_value={'<=1'}>}param_value))
)
If doesn't works please upload a sample to make tests.

Re: Set analysis
Satish Pawar Aug 6, 2015 2:25 AM (in response to Ruben Marin )Hi Ruben,
its not working.
here i'm attached sample QVF to test.
thanks in advance.

Community.qvf 5.0 MB

Re: Set analysis
Ruben Marin Aug 6, 2015 2:44 AM (in response to Satish Pawar)Hi Satish, this expression worked:
If(Count(total DISTINCT branch_type)>1,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_type),
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_name)
)
You can add more "levels":
If(Count(total DISTINCT branch_type)>1,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_type),
If(Count(total DISTINCT branch_name)>1,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_name),
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),loc_code),
))

Re: Set analysis
Satish Pawar Aug 6, 2015 3:19 AM (in response to Ruben Marin )Hi Ruben,
Thanks for your great efforts.
But have you checked it for loc_code and loc_type.
beacuse i think its not working for them.
i used following expression
If(Count(total DISTINCT branch_type)>1,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_type),
If(Count(total DISTINCT branch_name)>1,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_name),
If(Count(total DISTINCT loc_type)>1 ,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),loc_type),
If(Count(total DISTINCT loc_code)>1 ,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),loc_code)))))
please check

Re: Set analysis
Ruben Marin Aug 6, 2015 3:28 AM (in response to Satish Pawar)You only need the one I posted:
If(Count(total DISTINCT branch_type)>1,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_type),
If(Count(total DISTINCT branch_name)>1,
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),branch_name),
Count({<param_name={'TEMP'}, param_value={'<=1'}>} param_value)
/Aggr(Nodistinct count({<param_name={'TEMP'}, param_value={'<=1'}>}param_value),loc_code),
))
I don't see any selections where this doesn't work, but there are many possibilities to select in the chart, tell me with wich selections this doesn't works.

Re: Set analysis
Satish Pawar Aug 6, 2015 3:41 AM (in response to Ruben Marin )It is not showing any loc_type and when i replaced loc_code with loc_type in 3rd expression
it doesn't show any loc_type and directly shows loc_code.
and also that loc_code it shows below 100%.

Re: Set analysis
Ruben Marin Aug 6, 2015 3:50 AM (in response to Satish Pawar)In the qvf you posted, how can I see any loc_type? I don't see any loc_type in the dimensions, so if you replaced loc_code with loc_type in 3rd expression it's normal it shows below 100%.

Re: Set analysis
Satish Pawar Aug 6, 2015 4:08 AM (in response to Ruben Marin )oops...sorry its my mistake.
QVF which i given to you hadn't any loc_type in drilled group which i used in dimension.
sorry for taking time.

Re: Set analysis
Ruben Marin Aug 6, 2015 4:43 AM (in response to Satish Pawar)No worries, if you need working with loc_type upload a sample and I can help to make it work.
The login will be the same, just add another nested if with the next level of the group.









Re: Set analysis
Anand Chouhan Aug 3, 2015 7:09 AM (in response to Satish Pawar)Hi,
With your expression use the Dual function as well as in your expression that may solve your problem.
Regards,
Anand







