DAX Out of Memory

[andrearaujo]

Hello Everyone, 

Could you guys please help me?

I have a table with customers and products recommended for each customer and the date that the recommend was made, and I would like to calculate the recommended sales that I had for each customer and each product after the recommendation date.

But I'm getting out of memory. There is something wrong in my DAX, or there is another way to do this measure?

# Sales After Recommender = 

VAR recommender_date= MINX('recommendations', 'recomendations'[date])

VAR Customer_= SELECTEDVALUE('recomendations'[customer])

VAR Product_ = SELECTEDVALUE('recomendations'[product])

RETURN

CALCULATE([# sales], FILTER(Dates, Dates[FullDateAlternateKey] >= recommender_date), FILTER('dim_Customer', 'dim_Cliente Codigo'[CODIGO] = CLIENTE_), FILTER(dim_PID, dim_PID[PID] = PID_))

Thank you.

[v-robertq-msft]

Hi, @andrearaujo 

Have you followed the DAX formula posted by smpa01 to find the improvement to your memory in the desktop?

If so, would you like to mark his reply as a solution so that others can learn from it too?

For more info about the DAX formula improvement in the Power BI, please refer to these links:

https://powerbi.microsoft.com/fr-ch/blog/4-recommended-practices-for-new-dax-users/

https://maqsoftware.com/expertise/powerbi/dax-best-practices

https://blog.enterprisedna.co/dax-calculations-in-power-bi-best-practices/

Thanks in advance!

How to Get Your Question Answered Quickly 

Best Regards,

Community Support Team _Robert Qin

If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.

[smpa01 · Accepted solution]

@andrearaujo  try this and see if it improves

# Sales After Recommender = 

VAR recommender_date= MINX('recommendations', 'recomendations'[date])

VAR Customer_= SELECTEDVALUE('recomendations'[customer])

VAR Product_ = SELECTEDVALUE('recomendations'[product])

RETURN

CALCULATE([# sales],FILTER(
CROSSJOIN(VALUES(Dates[FullDateAlternateKey]),VALUES('dim_Cliente Codigo'[CODIGO]),VALUES(dim_PID[PID])),
Dates[FullDateAlternateKey] >= recommender_date&&'dim_Cliente Codigo'[CODIGO] = CLIENTE_&&dim_PID[PID] = PID_))

. I guess it can be optimized even further.