SQL Convert Rows to Columns and Columns to Rows without using Pivot Functions

[@ankitbansal6]

Master the art of SQL with my zero to hero SQL for data analytics course. www.namastesql.com/course-detail/think-sql-go-from-zero-to-hero-english

[@as-youtuber6215]

Man you might don't know how much you are helping other people. True gem

[@ankush_in_sync5998]

One of the unique thing Ankit sir does is giving the table script in the description. By this we can copy the script and practice. Thank you.

[@ankitbansal6]

Cheers 🥳

[@sanchitvaid7017]

I was asked this question in an interview (US based Health insurance company) but I only knew how to solve it using Pivot function and could not answer it. I wish if I had found this video sooner!

[@bhooshan25]

Practising from your videos before my interview. Thanks for the videos.

[@ankitbansal6]

Best of luck

[@RAAMASHAAMYC]

I have been working with SQL for a decade now. Now only I know that we can use 'val' for producing result instead of the field name itself. True Genius.

[@ankitbansal6]

val is a column in the table 😊

[@jececdept.9548]

​@@ankitbansal6 pls suggest how to study from so many videos..is there any pdf of these available.?

[@ankitbansal6]

@@jececdept.9548 just solve 1 question at a time everyday

[@lokeshvaishnav218]

Man you are an amazing person helping a lot of people

[@RajeshS-b6b]

🎯 Key Takeaways for quick navigation: 00:14 📊 Learn how to convert rows to columns and vice versa in SQL without using pivot functions. 01:25 🧮 Use SQL case statements to pivot data from rows to columns for specific components. 04:27 🧾 Practice using case statements with sum for efficient data pivoting in SQL. 05:28 🔄 Unpivot data by creating a new table using SELECT INTO and then using UNION to achieve the desired format. 08:18 📚 Mastering case when with sum is crucial for handling data transformations, and it's a valuable skill for interviews and job tasks. Made with HARPA AI

[@ratiranjannayak5712]

Thank you for your video. I have an exact requirement and your video makes it very clear with a proper description and makes it easy to understand. Thank you so much

[@ankitbansal6]

Glad it helped you 😊

[@Gaurav-e6i2e]

Thank you very much, Sir. The same question I have been asked in the interview. And because I had already watched this video, I could be able to solve this in the interview.

[@ankitbansal6]

Excellent

[@shubhamgoyal3358]

Good one ankit, I am watching daily 1-3 video from your channel to prepare for faang

[@ankitbansal6]

Awesome

[@brianducharme3418]

Sir you are brilliant, so much easier then using SQL pivot operator. Thank you kindly.

[@ankitbansal6]

Happy to help

[@sujaa1000]

One of the best channels, thank you very much Ankit!

[@ankitbansal6]

My pleasure

[@Exceltech810]

it's hard to see so so clear and easy understanding video. thank you

[@ankitbansal6]

Glad it was helpful!

[@CricketLivejavwad]

Thank you for the awesome video.I tried to solve this by myself after getting hint by this video and after struggling finally I was able to solve this problem and one more thing I want to tell you this concept is used in my job role also. Thanks again sir :)

[@ankitbansal6]

Glad it helped!

[@CricketLivejavwad]

Thank you sir ​@@ankitbansal6

[@girishggirishg1083]

I got a query related to this, thank you now i got some clarity.

[@ashutoshpandey1452]

this is what I was searching for.. thanks man

[@tupaiadhikari]

Thanks Ankit for the video, we have to use pivot function using pandas for subsequent ml work, but now we can do the data transformation directly in SQL and post which we can directly train the ml model

[@ankitbansal6]

Yes, you are right

[@sharu164]

Thanks Ankit for your time on this video 😊 appreciate all your effort

[@shoaibshaikh8909]

Hi ankit, thank you so much for simplifying the concepts. It would be great if you could make similar videos using the sum and case concept solving real interview questions or problems

[@ankitbansal6]

I have several videos on my channel using that concept. 😊

[@jaydeeppatidar4189]

Your videos has golden content! Just wowwwww!

[@ankitbansal6]

🙏🙏

[@shilpibiswas3110]

Hello ankit sir, Truly blessed to come across your superb videos....thank you so much sir and please keep helping is this way by more of ur interview questions on sql, pythons

[@ankitbansal6]

Keep watching

[@SachinKumarHS]

i have also tried this approach using window functions - select emp_id, salary, hike_percent, bonus from ( select emp_id, val as salary, lead(val) over a as hike_percent, lead(val,2) over a as bonus, row_number() over a as rn from emp_compensation window a as (partition by emp_id order by salary_component_type desc) ) t where rn=1 Thank you

[@nigarmutallimova8461]

This was an awesome tutorial! Very informative without the extra puff. Thank you so much!

[@ankitbansal6]

Glad it was helpful!

[@ankushjain4128]

keep doing what you are doing really it is very helpful.

[@MotivationToMoveForward]

It's really awesome video. I have learnt some great insights from this video

[@ankitbansal6]

Glad to hear that!

[@Ashu23200]

Hi Ankit I am learning a lot from you. Thank you, Thank you again. Please clear my doubt. Pasting my query below. I tried this method. but following both this method and aggregating both ways coming same output select emp_id, sum(case when salary_component_type = 'salary' then val end) as salary, sum(case when salary_component_type = 'bonus' then val end) as bonus, sum(case when salary_component_type = 'hike_percent' then val end) as hike_percent from emp_compensation group by emp_id; select emp_id, max(case when salary_component_type = 'salary' then val end) as salary, max(case when salary_component_type = 'bonus' then val end) as bonus, max(case when salary_component_type = 'hike_percent' then val end) as hike_percent from emp_compensation group by emp_id; how???

[@ankitbansal6]

There is only one not null value in each group. So max min avg all same ..

[@Ashu23200]

@@ankitbansal6 ok

[@karthikvb1293]

Really a great concept, explained very well.

[@thedatakid1270]

Very much helpful video. Ankit sir please upload such more complex query video

[@saurabhkhare3315]

Well explained... Please make some videos on path to join Amazon as a Business Intelligence Engineer or Data Engineer.

[@ankitbansal6]

Sure.

[@adreamer9999]

Great video! Excellent step-by-step explanation. Thank you.

[@ankitbansal6]

🙏

[@rakeshgopidi4066]

@Ankit - Awesome very well explained !

[@eel3720]

Great work Ankit so helpful thank you.

[@ankitbansal6]

Glad it helped

[@alex45688]

Thanks for sharing much value , a small request if you can make such more videos god bless you

[@ankitbansal6]

Sure 😊

[@alex45688]

@@ankitbansal6 People like you are a great help to the society and for new learners, we pray many people support this channel Will inform other people also to join this channel

[@kawshalchettri6614]

Hello Ankit, I bought multiple courses on SQL but finally started to understand by watching your videos. Thank you so much. I want to know where can I get more use cases to practice for case statements using sum and other aggregate functions.

[@ankitbansal6]

Check out the complex SQL problems playlist

[@milindzuge906]

Informative content ❣

[@georgehall2882]

This was great and well explained. Thanks

[@ankitbansal6]

Glad you enjoyed it!

[@aravindsatheesh4086]

Nice Video, thank you. I have a doubt. Is there any way that we can dynamically create these columns during pivoting instead of explicitly typing the column values in each sum(case) statement? If we have more unique values and more rows, it might be useful if possible.

[@ankitbansal6]

There is pivot function to do that. I will make a separate video.

[@taniyasaini6830]

Very well explained. I just have one question which may sound very basic. How to distinguish between pivoting and unpivoting? i always get confused which one is what

[@ankitbansal6]

Row to column is pivot and reverse in unpivot.

[@mallikarjunpatil7553]

Great session 👍

[@ankitbansal6]

Glad you enjoyed it

[@MrKingoverall]

Thank you so much sir for your help. 👍👍🙏

[@ankitbansal6]

Keep watching 💪

[@tomthomas6985]

Using PIVOT, we can solve it by: select emp_id,salary,bonus,hike_percent from (select emp_id,salary_component_type,val from emp_compensation) p PIVOT( max(val) for salary_component_type in ([salary],[bonus],[hike_percent]) )as pivot_table

[@suketshah7697]

Thanks very well explained...!

[@ankitbansal6]

Glad it was helpful!

[@ramakumarguntamadugu1299]

very useful 👌👌👍👍

[@rajakrishnan8561]

Hi Ankit, thank you sharing such a wonderful tip. this solves purpose.. But i have text value in "Val" column. when i use sum function, it shows error invalid digit value 'L'. any idea to group text values

[@ankitbansal6]

You can use string agg function ..or can use min , max

[@LalithammaBestavaripeta]

Thanks a lot sir for explaining in detail.

[@ankitbansal6]

You are most welcome

[@gritmanish]

Thanks for explanation. It was nice

[@subhojeetlearning]

Very nicely explained

[@ankitbansal6]

Keep watching😊

[@ashishdukare1313]

Thank you for your efforts : )

[@ankitbansal6]

My pleasure!

[@vinipowerbi]

Thank you Ankit for awesome explanation. Is there another video to practice this topic?

[@ankitbansal6]

There are many in this playlist Complex SQL Questions for Interview Preparation: kzbin.info/aero/PLBTZqjSKn0IeKBQDjLmzisazhqQy4iGkb

[@niveditasingh1176]

Hi Ankit, Thanks for creating this video . It was really helpful . Can you please create more examples around this.

[@ankitbansal6]

Sure

[@emiliogarza6446]

Really good explanation, great job :)

[@ankitbansal6]

Thank you 😊

[@hairavyadav6579]

Today i learn new thing in sql

[@hairavyadav6579]

And I also share on LinkedIn... Great explanation

[@NaveenKumar-ig6gn]

Hi Ankit, i am trying to run the below query in MySQL but getting error like undeclared variable at into emp_compensation_pivot Select emp_id ,SUM(case when salary_component_type = 'salary' then val end) as Salary ,SUM(case WHEN salary_component_type = 'bonus' then val end) as Bonus ,SUM(case when salary_component_type = 'hike_percent' then val end) as hike_percent into emp_compensation_pivot from emp_compensation group by emp_id; Can you please guide me

[@ankitbansal6]

Use create table as instead of into

[@yatinshekhar787]

@@ankitbansal6 sir used create table statement also but it is not working is their any other alternative?

[@mohit231]

Great video

[@ankitbansal6]

Glad you enjoyed it

[@mohit231]

Please make a video on using Pivot functions and merge statement

[@ankitbansal6]

Sure.

[@NaveenKumar-fq4sb]

super sir make a more interview quations.....

[@toshitachakraborty4184]

I am a big fan if you Ankit 💓

[@ankitbansal6]

Thank you 😊

[@adityaf17]

This is very helpful! Thanks

[@ankitbansal6]

Glad it was helpful!

[@monasanthosh9208]

Select Emp_id, Sum(Case When Salary_component_type="Salary" then Val end) As Salary, SUM(Case When Salary_component_type="Bonus" then Val end) As Bonus, SUM(Case When Salary_component_type="Hike_Percent" then Val end) As Hike Percent from Emp Group by Emp_id Order by EMp_id;

[@desflan06]

Thank you, this help me a lot.

[@ankitbansal6]

Glad it helped!

[@ankithashetty9133]

Incase if the rows are way to high how do we dynamically bring them as column names, is there a way to do it? It's not possible to give names individually . In my case I should bring dates as column names from the row.

[@kushwaha2k]

Hello, Do we do this with dates in place of numbers?

[@shubhiverma2571]

Thankyou so much for this video

[@harshitaggarwal3554]

Thanks a lot Ankit it was very helpful. Just one question in case we have multiple column like let's say 10-15 then we need to write 10-15 case statement from row to column and 10-15 union all in case od column to row or do we have any other approach

[@ankitbansal6]

Alternatively you can use pivot and unpivot functions.

[@rashmidutta7151]

Hi Ankit, very crisply explained. Can you explain how each row gets created in both the cases.I mean execution of the query. Thanks in advance!!

[@ujjwalvarshney3188]

select a.emp_id ,a. val as salary , b.val as bonus ,c.val as hike_percent from emp_compensation a join emp_compensation b on a.emp_id = b.emp_id join emp_compensation c on a.emp_id = c.emp_id where a.salary_component_type = 'salary' and b.salary_component_type ='bonus' and c.salary_component_type = 'hike_percent' order by 1

[@LivenLove]

Are case statements faster than pivot ?

[@ankitbansal6]

I would say you have the flexibility to give multiple case conditions based on the requirements.

[@LivenLove]

Thanks for replying so promptly. Do you have or planning to release any spark project based content?

[@sandycool1822]

Hi Ankit, if possible pls create video on how to do pivoting dynamically. If new entries comes in future it will automatically comes from rows to columns

[@ankitbansal6]

Okay

[@JosueLopezXIX]

Nice video! Thank you.

[@pavitrashailaja850]

Well explained video!

[@chronoray8786]

So fantastic. Thank you so much. :)

[@ankitbansal6]

You are so welcome!

[@seelamrk]

Thank you . What if I add one more type called Gender that has "Male" or "Female" - we cannot do SUM or any numerical aggregate like MIN/MAX in that case - so how do we solve that?

[@shahmaarbaba]

what happend when someone enter the new salary component type the query set was only the exitsting data when someone enter the new salary component type so every time i change the query it's not good please help me this like whenever some add new salary component type how to automatically add val in query

[@ankitbansal6]

You need to modify the query. Or wrote a dynamic SQL using pl SQL or t SQL

[@snipezu805]

Thanks friend very helpful

[@ankitbansal6]

Welcome

[@kundankumar5520]

True superhero👍

[@kanchans3556]

Is 'select into' syntax for creating new table worked in MySQL ? I got error

[@scrt1111]

Thank you valuable share

[@artofdata-]

will query work if we have millions of records ? we can not union every Time

[@hiyaborah9167]

Hello Ankit! This was really helpful, thank you so much for explaining it step by step. I came across a similar problem but have quite a few doubts, could you please help me understand how the same concept can be applied for the below table: NAME PROFESSION Samantha Doctor Julia Actor Maria Actor Meera Singer Ashley Professor Ketty Professor Christeen Professor Jane Actor Jenny Doctor Priya Singer Here is the query that I used: select case when profession = 'doctor' then person_name end as 'doctor', case when profession = 'actor' then person_name end as 'actor', case when profession = 'professor' then person_name end as 'professor', case when profession = 'singer' then person_name end as 'singer' from occupation How do I exclude the nulls is this case?

[@ankitbansal6]

You just need to create one case statement with multiple when ..that will do

[@vigneshnagaraj7137]

Hi Hiya, Please post your solution . It will be helpful

[@amazhobner]

I guess this is a hackerrank question, you can just group by and take min/max.

[@chaitanyagn]

Hello Hiya I tried to solve the same problem using the above mentioned method. The above example has emp-id, we can apply GROUP BY on this emp-id and get the desired result(remove the null values). However in the problem, we dont have any such column to perform GROUP BY on and eliminate the nulls. To overcome this we create a new column ‘grouper’. What grouper does is, it ranks the candidates for each profession. For ex: If there are 4 doctors, we give a rank to each doctor. Similarly for other professions. Then we perform GROUP BY on grouper column and get the desired result. Here’s the piece of query I have used. SET @r1=0, @r2=0, @r3=0, @r4=0; WITH tb1 AS (SELECT CASE WHEN Occupation = 'Doctor' THEN Name END AS Doctor, CASE WHEN Occupation = 'Professor' THEN Name END AS Professor, CASE WHEN Occupation = 'Singer' THEN Name END AS Singer, CASE WHEN Occupation = 'Actor' THEN Name END AS Actor, CASE WHEN Occupation = 'Doctor' THEN (@r1:=@r1+1) WHEN Occupation = 'Professor' THEN (@r2:=@r2+1) WHEN Occupation = 'Singer' THEN (@r3:=@r3+1) WHEN Occupation = 'Actor' THEN (@r4:=@r4+1) END AS grouper FROM OCCUPATIONS ORDER BY Name) SELECT MIN(Doctor) AS Doctor, MIN(Professor) AS Professor, MIN(Singer) AS Singer, MIN(Actor) AS Actor FROM tb1 GROUP BY grouper; Hope this helps !

[@sz6618]

@@ankitbansal6 It will not solve problem if we use so

[@poorvaagnihotri7104]

Hi Ankit, Thanks for this explaination. I was asked below question in LatentView Analytics Interview , can you share a optimal solution for that Input:- class Sub A Maths B English A Social B Chemistry A Physics B Statistics A Biology B Maths A Science B Social Result: Class Sub1 Sub2 Sub3 Sub4 Sub5 A Science Physics Social Maths Biology B English Social Maths Statistics Chemistry

[@ankitbansal6]

You can use a similar case when statements as explained in video. Just use max or min along with case when.

[@sujitnayak3072]

The same query how can we perform if the value has non numeric values.

[@deepakcl1449]

Thank you so much ! 🙂

[@ankitbansal6]

You're welcome 😊

[@TheEJAZ88]

can this process be used another query where numbers are not involved for pivot

[@KukuG28]

This method is dependent on number of columns. How to make this dynamic like pivot. Pivot automatically takes new column into account whereas this method will require to change query to add every new column. Please help.

[@navkarwebstudio7490]

How to pivot If all my columns having date. If sum will work in that case?

[@VARISHROCKS]

Well explained , just a request Please make a video on how to upload DB tables on this Software , it would be a great help from ur end !!

[@ankitbansal6]

You mean you want to load data from files in your system to SQL server right ?

[@VARISHROCKS]

@@ankitbansal6 YES

[@sanvinoth]

Hi Ankit - Thanks for Video. How we can use this if we have char/varchar instead of integers in the table

[@ankitbansal6]

Use Max min

[@arnabsarkar6159]

@ankit How to dynamically pick pivot rows to columns without using pivot function?

[@kartiks792]

Hi Ankit great video. Just one doubt can sum function work on string values in the case statement? Example just adding another columns as name and pan no.

[@ankitbansal6]

Yes it works on string.

[@kartiks792]

@@ankitbansal6 thank you for the clarification. 😊

[@the_cocktail_kudi]

What if there are 2 columns example val and expense_amount. So i want 2 colums from 1 row. Salary_val and salary_expense_amount. How to do that?

[@mvshop-wm1rn]

Hi Ankit, what do we use in case of phone numbers with hyphen instead of sum. Thanks

[@ankitbansal6]

Max or min

[@rutujakelkar2895]

Hi Ankit, Your videos are helpful to learn SQL in depth, do you have any github repository where you have hosted all your YT questions and solutions ? Or something in text/ blog format ?

[@ankitbansal6]

I haven't done that yet. Planning that.

[@surojeetghosh8389]

hello sir. thank you for sharing these concepts.its really helpful.sir im an aspiring data analyst and i really want to become good in sql.i know the basics and intermediate concepts.from where should i practice the hard concepts like this which you are teaching?once again thank u.

[@raghuramm3297]

Hi Ankit, if we have other salary component we have to alter the query . Can we dynamically generate the columns(for each Salary component)without using the pivot function.

[@hariarayakeel]

Thanks a lot Baha’i, very helpful video. I tried this in report which has 200 columns to unpivot and 5 inner joins, finally the script has 2300 lines now (each union has 15 lines). I have used inner joins on each union. Is it the a way to refine the script more!

[@the_class_apart]

This solved my potential career-ending problem, thanks a lot brother. How can I perform high performance on this when I am working on big data?

[@ankitbansal6]

It's optimised technique only.

[@CosmicKaBot]

My Answer: with pviot as (select emp_id,sum(case when salary_component_type='salary' then val else 0 end) as Salary, sum(case when salary_component_type='bonus' then val else 0 end) as bonus, sum(case when salary_component_type='hike_percent' then val else 0 end) as hike_percent from emp_compensation group by 1) select emp_id,Case when Salary then 'Salary' end as salary_component_type, case when salary then Salary end as val from pviot union select emp_id,Case when bonus then 'bonus' end as salary_component_type, case when bonus then bonus end as val from pviot union select emp_id,Case when hike_percent then 'hike_percent' end as salary_component_type, case when hike_percent then hike_percent end as val from pviot;

[@as-youtuber6215]

Man you might don't know how much you are helping other people. True gem

[@ankitbansal6]

🙏

[@priyachauhan813]

Thanks for your video, but what if I want to unpivot 35 different months columns into row based and 35 quantity columns for the corresponding month into row based. This method is not useful fod that, also the unpivot function is not efficient here, query processing times is going more then an hour, could you please suggest any approach for this

[@ankitbansal6]

You have only 2 options. Either use pivot unpivot functions or do like in video.