Calculation for prediction

[#1]

Hello,

I need help regarding prediction values. I describe my case briefly :

I have some TestObject and their Estimatedvalue are calculated. then from the value they can be sell in one or two steps with estimated efficiency.

Then I want to have the predicted sale in future by sum (Estimatedsale*EstimatedValue) in each month from different TestObject and plot them.

TestObject	EstimatedValue	Estimatedsale1	EffectMonth	Estimatedsale2	EffectMonth
A	7,2	0,8	2017-04	0,2	2017-07
B	6	0,3	2017-06	0,6	2017-09
C	8,1	0,5	2017-09	0,4
D	8,5	0,7	2017-05	0,2
E	4	0,6	2017-06	0,4	2017-09
F	5,7	0,5	2017-06	0,3	2017-07
G	7,8	0,4	2017-09	0,4

[Anil_Babu_Samineni]

In theoretically, I got understand. when practical comes i need expected result as well. Is that possible to show expected outcome?

[swuehl]

You could transform your data in a table that only shows a single effect month:

LOAD TestObject,

    EstimatedValue,

    Estimatedsale1 as Estimatedsale,

    Date#(EffectMonth,'YYYY-MM') as EffectMonth,

    '1' as Step

FROM

[https://community.qlik.com/thread/261218]

(html, codepage is 1252, embedded labels, table is @1);

CONCATENATE

LOAD TestObject,

    EstimatedValue,

    '2' as Step,

    Estimatedsale2 as Estimatedsale,

      Date#(EffectMonth1,'YYYY-MM') as EffectMonth

FROM

[https://community.qlik.com/thread/261218]

(html, codepage is 1252, embedded labels, table is @1);

Then you can use this EffectMonth as dimension: