Optimizing DAX queries in Power BI - Cheat Sheet & Quick Revision

Recall & Review

beginner

What is the main goal of optimizing DAX queries?

To make DAX calculations run faster and use less memory, improving report performance and user experience.

intermediate

Why should you avoid using 'FILTER' on large tables in DAX?

Because 'FILTER' scans the entire table row by row, which can slow down queries on big datasets.

beginner

How does using variables (VAR) help optimize DAX queries?

Variables store intermediate results so calculations are done once, reducing repeated work and speeding up queries.

intermediate

What is the benefit of using CALCULATE with simple filters instead of complex FILTER expressions?

CALCULATE with simple filters lets the engine use optimized storage indexes, making queries faster than complex row-by-row filters.

beginner

Why is it important to reduce the number of columns in your data model for DAX performance?

Fewer columns mean less data to scan and store, which speeds up calculations and reduces memory use.

Which DAX function is generally faster for filtering large tables?

A CALCULATE with simple filters

B FILTER with complex conditions

C SUMX over entire table

D ALL to remove filters

CALCULATE with simple filters uses optimized engine paths, making it faster than FILTER on large tables.

What does using VAR in a DAX measure help you avoid?

A Data refresh

B Changing data types

C Data import errors

D Repeated calculations

VAR stores a value once so it can be reused, avoiding repeated calculations.

Which practice improves DAX query speed the most?

A Reducing columns in the data model

B Adding more calculated columns

C Using many nested FILTER functions

D Using complex IF statements

Reducing columns reduces data size and speeds up calculations.

What is a downside of using FILTER on large tables?

A It disables relationships

B It scans every row, slowing performance

C It deletes data

D It changes data types

FILTER evaluates each row, which can be slow on big tables.

How can you check which DAX queries are slow in Power BI?

A Restart Power BI Desktop

B Use Excel formulas

C Use Performance Analyzer tool

D Change report theme

Performance Analyzer shows query times and helps identify slow DAX.

Explain three ways to optimize DAX queries for better performance.

Think about filtering, reusing calculations, and data size.

Describe why FILTER can slow down DAX queries and how to avoid it.

Consider how row-by-row operations affect speed.

Practice

1.

What is a simple way to improve the performance of a DAX query in Power BI?

SUM(Sales[Amount]) is slow. What should you do?

easy

A. Avoid using filters in the query

B. Add more columns to the table

C. Use complex nested IF statements

D. Use variables to store intermediate results

Solution

1. Step 1: Understand the role of variables in DAX

   Variables store intermediate results to avoid repeated calculations, making queries faster.

2. Step 2: Apply variables to the example

   Using variables in SUM(Sales[Amount]) can reduce repeated scans and improve speed.

3. Final Answer:

   Use variables to store intermediate results -> Option D

4. Quick Check:

   Variables improve performance [OK]

Hint: Use variables to avoid repeated calculations [OK]

Common Mistakes:

• Adding unnecessary columns slows down queries
• Using complex nested IFs increases calculation time
• Ignoring filters can cause large data scans

2.

Which of the following is the correct syntax to define a variable in a DAX measure?

Measure = ?

easy

A. Measure = DEFINE Total = SUM(Sales[Amount]) USE Total

B. Measure = VAR Total = SUM(Sales[Amount]) RETURN Total

C. Measure = LET Total = SUM(Sales[Amount]) THEN Total

D. Measure = SET Total = SUM(Sales[Amount]) OUTPUT Total

Solution

1. Step 1: Recall DAX variable syntax

   DAX uses VAR to declare variables and RETURN to output the result.

2. Step 2: Match syntax with options

   Measure = VAR Total = SUM(Sales[Amount]) RETURN Total correctly uses VAR Total = SUM(Sales[Amount]) RETURN Total.

3. Final Answer:

   Measure = VAR Total = SUM(Sales[Amount]) RETURN Total -> Option B

4. Quick Check:

   VAR and RETURN are correct DAX keywords [OK]

Hint: Remember: VAR ... RETURN in DAX for variables [OK]

Common Mistakes:

• Using LET or DEFINE instead of VAR
• Forgetting RETURN keyword
• Using SET or OUTPUT which are not DAX keywords

3.

Consider this DAX measure:

Measure =
VAR FilteredSales = FILTER(Sales, Sales[Region] = "West")
RETURN SUMX(FilteredSales, Sales[Amount])

What is the main performance issue with this measure?

medium

A. FILTER creates a filtered table and SUMX iterates over it, which can be slow if Sales is large

B. FILTER is applied before SUMX, which is efficient

C. FILTER creates a filtered table but SUMX uses the original table, causing repeated scans

D. FILTER creates a table but SUMX iterates over original Sales table

Solution

1. Step 1: Analyze FILTER and SUMX behavior

   FILTER creates a filtered table of Sales where Region is "West". SUMX iterates over this filtered table.

2. Step 2: Understand performance impact

   If Sales is large, creating and iterating over a filtered table can be slow. Better to filter early or use CALCULATE with SUM.

3. Final Answer:

   FILTER creates a filtered table and SUMX iterates over it, which can be slow if Sales is large -> Option A

4. Quick Check:

   FILTER + SUMX on large tables can slow queries [OK]

Hint: Avoid FILTER + SUMX on big tables; use CALCULATE [OK]

Common Mistakes:

• Thinking SUMX iterates original table, not filtered
• Assuming FILTER before SUMX is always fast
• Ignoring table size impact on performance

4.

Identify the error in this DAX measure that aims to optimize performance:

Measure =
VAR TotalSales = SUM(Sales[Amount])
RETURN
CALCULATE(TotalSales, FILTER(Sales, Sales[Region] = "East"))

medium

A. Using FILTER inside CALCULATE is always slow

B. TotalSales should be inside FILTER function

C. TotalSales is a scalar evaluated outside CALCULATE, so the filter does not apply

D. No error; measure is optimized

Solution

1. Step 1: Check variable type and usage

   TotalSales = SUM(Sales[Amount]) is a scalar value computed in the original filter context.

2. Step 2: Analyze CALCULATE arguments

   CALCULATE(TotalSales, FILTER(Sales, Sales[Region]="East")) is syntactically valid, but since TotalSales is a pre-computed scalar, the modified filter context does not re-evaluate it. It returns the unfiltered total instead of East sales, defeating the performance optimization goal for filtered calculation.

3. Final Answer:

   TotalSales is a scalar evaluated outside CALCULATE, so the filter does not apply -> Option C

4. Quick Check:

   Scalar variable used incorrectly with CALCULATE filter [OK]

Hint: Scalar variables computed outside aren't affected by CALCULATE filters [OK]

Common Mistakes:

• Thinking CALCULATE re-evaluates scalar variables in new context
• Thinking FILTER inside CALCULATE is always slow
• Not distinguishing scalar vs table types

5.

You want to optimize a DAX measure that calculates total sales for the current year only. Which approach is best for performance?

Option 1:
Measure = CALCULATE(SUM(Sales[Amount]), FILTER(Sales, YEAR(Sales[Date]) = YEAR(TODAY())))

Option 2:
Measure = CALCULATE(SUM(Sales[Amount]), Sales[Date].[Year] = YEAR(TODAY()))

hard

A. Option 2 is better because it uses column filter, not row filter

B. Option 1 is better because FILTER is more flexible

C. Both options have same performance

D. Neither option works; use SUMX instead

Solution

1. Step 1: Compare FILTER vs direct column filter

   Option 1 uses FILTER which scans row by row, applying YEAR function, which is slow.

2. Step 2: Understand column filter advantage

   Option 2 uses a direct filter on the date column's Year attribute, which is faster because it uses existing column indexing and avoids row-by-row evaluation.

3. Final Answer:

   Option 2 is better because it uses column filter, not row filter -> Option A

4. Quick Check:

   Use column filters over FILTER for better speed [OK]

Hint: Filter on columns, not rows, for faster DAX queries [OK]

Common Mistakes:

• Using FILTER with row context unnecessarily
• Assuming FILTER is always faster
• Ignoring built-in column filters