Future Simulation without existing data, based on past data?

[#1]

Hi,

I have e.g. turnover data for the last years. The calendar information is hold in a calendar table, going from e.g. 2000 to 2020. I would like to show in a diagram the turnover for the last years by period, AND how the turnover will develop in the next year based on the development of the current year. For the next year I do not have data.

I thought about using the calendar table for the dimension, and splitting the formula at today. For all dates below today I use the information I have, for all future dates, I calculate based on the values of this year. For example turnover growed by 2% from January 2011 to February 2011, by 3% the next month, and so on. So how does it develop for January 2012, based on the turnover in December 2011, growing by the calculated 2%.

When I try that, I am told that there are no information available... Do you have an idea or example for that?

Thanks a lot,

Martin

[#2]

Hi,

did indeed no one ever try this?

Martin

[rohit214]

hi

if i understood you properly

then try this

use an in put box add new variable (forecast%)

abd in your chart expresion use

Sum ([turnover.])*(1+forecast%/100)

thanks

rohit

[#4]

Hi,

I attached an example.

For example I have sales-values for 2011 (and before). I do NOT have any values for 2012. I just generated the calendar-table including 2012.

In the example the upper diagram shows sales for 2011. Based on these values I want something like the lower diagram (created in Excel).

This one contains simulated values for 2012, based on the experience for 2011. e.g. since I know that sales increased by 154,- from dec.2010 to jan.2011, i want to estimate that the value for jan.2012 will be  dec.2011 + ( (jan.2011-dec.2010) * (1 + est.plus) ). And so on.

Was that understandable?

Martin

[rohit214]

hi i undersyood ur point

try this

sum({<[Calendar Month Year]={$(=max([Calendar Month Year])}>}Sales)

+

((sum((<[Calendar Month Year]={$(=min([Calendar Month Year])}>}Sales)

-

sum((<[Calendar Month Year]={$(=min([Calendar Month Year])-1}>}Sales)

*

(1+forecast/100))

hope this will help you

thanks

rohit

[#6]

Hi,

thanks for your response. I tried

=sum({<[Calendar Month Year]={$(=max([Calendar Month Year]))}>}Sales)

+

((sum({<[Calendar Month Year]={$(=min([Calendar Month Year]))}>}Sales)

-

sum({<[Calendar Month Year]={$(=min([Calendar Month Year])-1)}>}Sales))

*

(1+forecast/100))

but the values are not only empty, I stil don't get dimension values for the 2012 periods.

Thanks,

Martin

[rohit214]

hi

i think for getting value of you have to change your system date

your script is running  acccording to your system date

may be i am wrong

thanks

rohit

[#8]

Hi rohit,

thanks for your answer, unfortunately I don't understand what you mean...

Cheers,

Martin

[#9]

Hi rohit,

thanks for your last response (which disappeared somehow). If I got it right, it is actually not possible to create a simulation of data based on not-yet-existing date values, right?

Regards,

Martin

[rohit214]

hi Martin

wish you a very very happy new year,

as i discussed about your problem with other then i came to know

Forecasting is the toughest part in qlik view now  as per my knowledge.

According to your problem

that forecasting date should be in your database otherwise you can not able to create chart or whatever you want

regards

rohit