Recently, I interviewed a few people for Power BI roles and here are some questions I asked in the initial rounds to assess their skill level. I’ve included sample answers below so you know how to approach such questions.
Prefer video? See this...
I made a video with 10 questions and my (elaborate) answers to them. Watch to learn how to answer such interview questions. See it below or on my YouTube channel.
1. What are the main differences between Power BI Desktop and Power BI Service?
Answer cues:
- Power BI Desktop is the main tool for creating or authoring reports.
- We use PBI desktop to clean up data, create data model, set up measures, build and refine charts and construct the reports.
- We then use Power BI Service to share the reports with the audience.
- Desktop = creation tool, Service = consumption tool
2. Explain the difference between a measure and a calculated column?
- Measures:
- calculate things on top of tables.
- calculated in the run-time based on the evaluation context.
- are usually aggregates.
- example: Shipment Count = COUNTROWS(Shipments)
- Calculated columns:
- are part of the table.
- are calculated at the time of creation based on the row-context.
- have one value per row.
- example: Discount Rate Row-wise calculation of discount rate based on order quantity
- Both measures and calculated columns use the same DAX language.
3. Describe the steps you would take to connect multiple data sources in a Power BI report?
- We can use Power Query to connect more than one source of data.
- We can also use PQ to merge / append / combine the data as needed.
- We can selectively load the data to Power BI and keep the rest in PQ level but customizing the load behaviour.
4. What do 1, * and arrow (â–¶) mean in the data model diagram?
- In a typical star-schema, we will have one to many relationship between dimension and fact table.
- The one side is denoted with a 1 and many side is denoted with a *
- The arrow indicates the direction of filter propagation.
- Power BI allows us to customize the filter direction (uni or bi-directional) and also have many-to-many relationships.
5. What is DAX and give me an example of DAX you've recently used?
- DAX stands for Data Analysis eXpressions. It is the main language of Power Pivot.
- Give examples based on your experiences and tell why / how they helped you solve problems.
6. How do you optimize the performance of a slow Power BI report?
- Answer questions like this based on your experience. If you have never optimized something, be honest and say that. Then give theoretical answers.Â
- Key optimization startegies:
- Reducing the data (filtering at PQ, Source level)
- Removing unwanted visuals, interactions and bookmarks
- Using performance profiler (Optimize ribbon in Power BI) to measure the performance of a page and identifying the problem areas.
- Setting up aggregate tables, pre-calculated views.
- [NEW] Using visual calculations instead of measures.
- Upgrading the Power BI on-prem servers, database servers (for direct query)
- Running time-consuming operations (such as PQ transformations) at low usage times (midnights, evenings)
7. How do you publish a Power BI Report to the service?
- Save and test the report. Make sure all calculations are correct and visuals represent the truth.
- Test any connections, refresh processes too.
- Publish the report (Publish button) and set the correct workspace.
- Test the report on the service view and make sure right people have access to the report.
- Optional, Send an email or share the report with the audience.
8. Explain RLS (Row Level Security) in Power BI?
- Row Level Security allows us to provide access to the relevant data to right people.
- For example, we can use RLS to show only USA data to the USA regional manager.
- We can use “roles” option in Power BI to set and test the roles.Â
9. What is a Slowly Changing Dimension (SCD) and how do you handle it in Power BI?
- SCD (Slowly Changing Dimension) is a dimension (or aspect of a dimension) that changes slowly over time.
- Give examples from your industry or previous work.
- Example: In our product table, we have a feature called cocoa percentage. For certain products this is changed once in a while based on customer feedback. This is an example of SCD.
- We can either replace the old values with new ones or create effective date based records.Â
10. How would you handle missing data?
- We can use either Power Query or Power Pivot to handle missing values in our data.Â
- Power Query shows missing values in a column thru Column Quality feature. We can use this to identify and deal with missing values.
- The key strategies for dealing with missing data are:
- Removing missing values
- Going back to source and fixing the problem
- Replacing missing values with an approximation (imputation)
- Ignoring missing values
Need more help? Watch the video
I made a video with 10 questions and my (elaborate) answers to them. Watch to learn how to answer such interview questions. See it below or on my YouTube channel.
Learn more: Power BI Weekend (2023)
My annual Power BI event – Power BI Weekend is happening this year on November 18 & 19. In this 4 hour event (2 hours on Saturday – Nov 18 & 2 more on Sunday – Nov 19), you will learn the Power BI Essentials to say yes to your next challenge.
The topics covered are,
- What is Power BI? How to use it?
- Power BI vs Power Query vs. Power Pivot
- Understanding Power BI Visual interactions & customizing them
- Data cleanup and transformations with PQ
- Setting up a star schema data model
- Creating and using DAX measures
- Interactive storytelling in Power BI
- Saving & publishing your work
- Resources to learn more
- Q&A with you
Tickets are on sale now. Book yours before they sell-out.
15 Responses to “Compare 2 Lists Visually and Highlight Matches”
Hi,
I solved this in a little different way.
We have 2 lists, one starts at A1 and other at B1, both are vertical arrays.
First thing is define 2 named ranges, list1 and list2:
list1 refers to "=OFFSET(Sheet1!$A$1;0;0;SUMPRODUCT(--(Sheet1!$A$1:$A$1000""));1)"
list2 refers to "=OFFSET(Sheet1!$A$1;0;0;SUMPRODUCT(--(Sheet1!$B$1:$B$1000""));1)"
this way lists will be dynamically sized when you had or remove elements (you can't have blanks and you can't have more than 1000 elements).
Then I use conditional formatting in column A when this formula is true:
"=NOT(ISERROR(MATCH(A1;list2;0)))"
and "=NOT(ISERROR(MATCH(B1;list1;0)))" to list2.
This way we eliminate the need for auxiliary columns or lists.
Hope you like my way! 😀
Nunes
Simple conditional formatting formula.
Assuming lists vertical lists starting in A1 & B1
To highlight just one column (assume B for example)
Conditional formatting>New Rule>by formula
=MATCH(B1,$A$1:$A$99,0)
Set the cell fill to what ever color you prefer & press OK
To highlight both columns repeat with this formula for cell in column A
=MATCH(A1,$B$1:$B$99,0)
This approach doesn't require named fields or addtl columns
glw
Say I had 1 list in A2:A20 and another in B2:B20.
To format all the items in column A that are repeated in column B I would use the following Conditional Formatting rule.
=IF(ISNA(VLOOKUP(A2,$B$2:$B$20,1,false)),true,false)
All the duplicates are highlighted. It us a very simple example of comparison.
I may be missing something here, but I usually highlight both my lists by holding ctrl eg A1:A20 E10:E40 then choose conditional formatting from the ribbon and then highlight duplicates, and this does it?
Lee, I was perplexed as well. I do the same thing you do with the conditional formating. A drag and click to highlight range and choose highlight duplicates does the trick for me.
I believe these methods are to check if an item from one list also appears in the other list. So if an item mentioned many times in one list if also mentioned in the other list or not.
The Conditional Formatting highlight duplicates feature will do this, but it will also highlight an item if it appears multiple times in the one column or list.
Hi, I would just like to know (if you are willing to share) which image editing program you use to make your image like above, like they are torn apart from bottom? I've been looking for long.
@i48998
Chandoo is on Holidays, but Chandoo uses Paint.Net
Paint.net is a free download available at http://www.paint.net/
.
I use CorelDraw/PhotoPaint
.
We both use the Snipping Tool (a freebe with Win Vista/10)
.
We both use Camtasia for doing screen captures to make animated GIFs where you see animation.
Here is how I would accomplish
(1) Define Names: List_1, List_2
(2) =ISNA(MATCH(D4,List_2,0))-1 (Conditional Format formula List_1)
(3) =ISNA(MATCH(D4,List_1,0))-1 (Conditional Format formula List_2)
ISNA will return 1 if NO Match and O if Match by adding a -1 will make: NO Match 0 and Match a -1 which is True
Hi all
this my first Post here
i think we can take Unique List for tow list to know what is not Duplicate By this Array formula
=IFERROR(INDEX($D$6:$D$33,SMALL(IF(ISERROR(MATCH($D$6:$D$33,$B$6:$B$33,0)),ROW($D$6:$D$33)-ROW($D$6)+1),ROWS($J$5:J5))),"")
and this one for Duplicate Value
=IFERROR(INDEX($D$6:$D$33,SMALL(IF(ISNUMBER(MATCH($D$6:$D$33,$B$6:$B$33,0)),ROW($D$6:$D$33)-ROW($D$6)+1),ROWS($J$5:J5))),"")
Don't forget to Enter This Formula by Pressing Ctrl+Shift+Enter
without wanting to ruthlessly self promote here, I do have an addin that does neatly compare two ranges, not just in columns, so you might want to check that out.
Having said that this is a pretty neat solution if you dont want to be going down the VBA or purchase route. I like it
however, could you not do something with the remove duplicates feature in Excel 2010 and then compare the resulting data set?
Hi, Chandoo! I've found yesterday your Excel website... What can I say? It's just awesome, Excellent. Being a developer for 30 years, more than 15 with Office products, and wow!, how many things I discovered in a couple of hours, and what pretty resolved.
I decided to take the long path of the newbies and read all your examples and write down by myself all of them, and when I arrived to this (the comparison of two lists) I think I've found a problem:
a) in "Step 4: Apply conditional formatting to Second List - Use the same logic, but this time the rule becomes =COUNTIF(count1s,$H6)" it should say "Step 4: Apply conditional formatting to Second List - Use the same logic, but this time the rule becomes =COUNTIF(count1s,$H6)>0", but this is a typing error that I believe all of us here might have discovered and corrected
b) the very problem: I wrote down two different lists, in different ranges, and with different number of elements, I specified the equivalent conditional formats, et non voilá!, I didn't get what expected. So I downloaded your example book, I checked range names, formulaes, conditional formats and all OK. So I copied -just values- from my book to yours, and I still couldn't achieve the goal.
I'm using Excel 2010 in spanish, I'm from Buenos Aires (Argentina), and my book is at your disposition whenever you considerate it appropiate.
Thanks in advance for your time, and again my congratulations for your work here.
Best regards.
SirJB7
Comparison of 2 lists visually with highlights
Author: SirJB7 / Date: 11-Dic-2011
Pros: no duplicated tables, no matrix formulaes, no named ranges, no VBA code, just conditional formatting
Cons: not found yet, comments and observations welcome
Features:
a) standard problem: highlights in orange/yellow elements existing in the other list
b) optimized problem: idem a) plus highlights in red/violet first occurrence of elements existing in the other list
Sheet contents:
a) conditional format, 1 rule per list (2 methods used)
A1:A20, first list
B1:B20, second list
a1) range A1:A20, condition =NO(ESERROR(BUSCARV(A1;B$1:B$20;1;FALSO))), format Orange ---> in english: =NOT(ISERROR(VLOOKUP(A1,B$1:B$20,1,FALSE)))
a2) range B1:B20, condition =CONTAR.SI(A$1:A$20;B1)>0, format Yellow ---> in english: =COUNTIF(A$1:A$20,B1)>0
b) conditional format, 2 rules per list (2 methods used)
D1:D20, first list
E1:E20, second list
b1) range E1:E20, condition 1 =Y(NO(ESERROR(BUSCARV(D1;E$1:E$20;1;FALSO)));COINCIDIR(D1;D$1:D$20;0)=FILA(D1)), format Red ---> in english: =AND(NOT(ISERROR(VLOOKUP(D1,E$1:E$20,1,FALSE))),MATCH(D1,D$1:D$20,0)=ROW(D1))
same range, condition 2 and format 2, same as a1)
b2) range E1:E20, condition =Y(CONTAR.SI(D$1:D$20;E1)>0;COINCIDIR(E1;E$1:E$20;0)=FILA(E1)), format Violet ---> in english: =AND(COUNTIF(D$1:D$20,E1)>0,MATCH(E1,E$1:E$20,0)=ROW(E1))
same range, condition 2 and format 2, same as a2)
Personally I like the a2) and b2) solutions, I think the formulaes are prettier.
I still don't know the rules of this website and forum, but it any precept is infringed I'm willing to share the workbook with the solution. If it breaks a rule, I apologize and promise that won't happen again.
Best regards for all!
Dear All i have a complicated situation...
1. I have two sheets of data Sheet1 and Sheet2Â (from various sources) - Both of these contain data matching and Not matching as well..
2. Now for me i need to build an excel where in i need to get sheet 3 with values that are present in a column of Sheet 1.
What ever Sheet 1 doesn't have i dont want those rows from sheet 2 to be populated into Sheet3.
Can any one help me out.
Hi Team
The above example is to compare partial name from 2 different columns.
If I want to cross check it in a single column. I have both correct and partial correct/match entries in a column. Is there any way I can find both the entries in the column.
Regards