Bring the current data considering changes their valid values over time


Hi Sam and guys.

I would like your help to solve a problem that I´m facing on my modeling approach

A same “contract” could change its values acording some considerations. So I have an specific data with valid values between an initial date and an end date.

I´d like to build a table where I could bring for any date the valid data in this specific date, so that I could apply all the time inteligence paterns and treat my model as an static one.

What I´ve already done:

I construct a sample model with two date tables and I can select the values filtering in different periods through 2 date slices.once for each table.

Margem = 

PnL_Data_Filter_Ini = 
 VAR vData = MAX('Calendar_Ini'[Date])
 CALCULATE([Margem];FILTER(Contratos;Contratos[DfimVig]>=vData && Contratos[DiniVig]<=vData))

PnL_Data_Filter_Fim = 
 VAR vData = MAX('Calendar_Fim'[Date])
 CALCULATE([Margem];FILTER(Contratos;Contratos[DfimVig]>=vData && Contratos[DiniVig]<=vData))

Pnl Delta = [PnL_Data_Filter_Fim]-[PnL_Data_Filter_Ini]

so I can calculate the results for different dates and compare them.

my problem is that as I need use an waterfall chart to explain the changes through time I would like to build an unique table where I can bring the valid data in each date.

So for this I imagine that I have to put in one table the valid values and consider just one date table.

How can I construct a table with this logic ?

Is my approach feasible?

Thanks in advance!

Best regards,

Giovanni Campolina


I’m not 100% understanding ever aspect of this.

One thing though I wouldn’t do it have two date tables. I don’t believe you need these. I rarely use two as you can solve nearly everything you need with one date table and just the correct DAX formulas.

The key to using one date table is to have one active relationship then the other is an inactive relationship. You can turn on the inactive relationship using a function called USERELATIONSHIP.

I showcase how to do this here

You should be able to then use a date slicer. That would give you a start and end date. Just the MIN and MAX of any context you select.

You could capture theses the same way in variable like you have done already.

From here I’m a little lost on what you need as I’m looking at all the data and don’t really know where to start.

Maybe once you have this setup lets focus on one specific calculation. Once you know how to solve one of these, you’ll likely understand how to solve the rest.



Thank you Sam.

I will try to apply the approach that you indicated.

Best regards,