Accumulating values on each time range

[#1]

Hello!

Well i think this should be easy but the worse thing is that i can not solve it yet.

I have a table containing an order list and time span. I am having a chart viewing the value for each month which is working fine. Now i want to create a chart to list the accumulated value from the previous month/months on each month. That is to say jan will show only jan values but feb with show the sum of values from jan and feb, and so on at last dec will show the sum of the values from jan to dec.

How can i achieve this in the easiest way ?

[#2]

Hi,

In your graph properties and in the Expression tab , try to check the full accumulation (bottom left ).

Philippe

[#3]

Hi,

try to check "Full Accumulation" in Expression tab.

I hope this will work 🙂

Milda

[#4]

Thanks for your answer.

Sorry but i already tried that and it does not work. Full accumulation accumulates the values of the expression (y-axis) and does not help my % expression (X/Y*100) as i need to accumulate both the X values and the Y values before dividing other wise it will be wrong in the % numbers. Below is one of my expressions that is working for values for each month only. I want to modify this to use it for accumulated values in the expression:

= Count(distinct if(type = 'X', wo)) / Count(distinct wo)) * 100

[#5]

May be i did not explain my problem correctly from the beggining.

I need to accumulate the value in my expression before deviding and not after.

[#6]

Do you have an exemple of your application ?

You can probably do what you want to achieve using the Set Analysis Syntax.

To help you , we need more information . The application could be a good start.

Philippe

[johnw]

Here's one way to accumulate data. If it's conditional, like a YTD accumulation for multiple years, you can add an if() statement around the above():

rangesum(above(ExpressionName),Expression)

So for your expression, I think you'd want this:

rangesum(above(FirstExpressionName) ,count({<type={'X'}>} distinct wo))
/ rangesum(above(SecondExpressionName),count( distinct wo))

And then display it as a percentage using the Number tab instead of multiplying by 100. I used set analysis for the type='X' check because it's usually faster, not because it's necessary here.

[#8]

Hello Philippe and all !

here comes an example file.

[johnw · Accepted solution]

OK, here's what I think you're asking for, using basically the technique above.

[#10]

Thanks alot it really works as i wanted