# Expand JSON dataset into columns instead of rows **URL:** \ **Category:** Ask a question\ **Tags:** Building-Flows\ **Created:** [August 19, 2022, 6:21pm UTC](https://community.parabola.io/t/expand-json-dataset-into-columns-instead-of-rows/2479 "2022-08-19T18:21:51Z")\ **Posts on this page:** 6\ **Page:** 1
**Author:** ![Axel\_Delgado](https://avatars.discourse-cdn.com/v4/letter/a/ea5d25/32.png) [@Axel\_Delgado](https://community.parabola.io/u/Axel_Delgado)\ **Post date:** [August 19, 2022, 6:21pm UTC](https://community.parabola.io/t/expand-json-dataset-into-columns-instead-of-rows/2479/1 "2022-08-19T18:21:51Z")
I’m pulling from an API to send to Google Sheets, but there is some deep nested data in the JSON where I only need some specific values out of there. However, I want to maintain the amount of rows I have, and not have them multiplied based on the nesting levels. Having extra rows means I need to find a way to merge them to avoid losing data. If I can flatten them into extra columns, I can just delete the extra columns, but always keep all the values linked to each row. I had a search but couldn’t find a related request. Many thanks. Just discovered Parabola after 4 days of fighting to find a tool that can do this properly. Parabola perfectly fits the bill, and I’m SO close now! ---
**Author:** ![Emory\_Stainbrook](https://sea2.discourse-cdn.com/flex020/user_avatar/community.parabola.io/emory_stainbrook/32/1772_2.png) [@Emory\_Stainbrook](https://community.parabola.io/u/Emory_Stainbrook)\ **Post date:** [August 19, 2022, 10:49pm UTC](https://community.parabola.io/t/expand-json-dataset-into-columns-instead-of-rows/2479/2 "2022-08-19T22:49:58Z")
Hi @Axel_Delgado- Welcome to the Parabola Community! Happy to help as you build this out. If the duplicate rows have overlapping information, it sounds like you may want to use a [Merge duplicate rows step](https://parabola.io/transform/merge-duplicate-rows). This would merge the information from those rows into one. Once you’ve done that, you would then expand that information into multiple columns using a [Split column step](https://parabola.io/transform/split-column). Let me know if you have any questions! ---
**Author:** ![Axel\_Delgado](https://avatars.discourse-cdn.com/v4/letter/a/ea5d25/32.png) [@Axel\_Delgado](https://community.parabola.io/u/Axel_Delgado)\ **Post date:** [August 20, 2022, 7:46am UTC](https://community.parabola.io/t/expand-json-dataset-into-columns-instead-of-rows/2479/3 "2022-08-20T07:46:13Z")
Thanks @Emory_Stainbrook. I was asking to do it directly because other tools have the option to expand directly into a single row during extra columns. I understand then that there’s no way around it and I have to do that process, correct? I’m trying to do what you suggest, but I’m struggling. ![image](https://us1.discourse-cdn.com/flex020/uploads/parabola/original/2X/6/60f6bccd2d8e54d0823db96378012855257d3c14.png) The merge duplicate rows is the conflicting step and can’t find a way of doing what I want. Here’s the data structure before merging. ![image](https://us1.discourse-cdn.com/flex020/uploads/parabola/original/2X/d/dec3f42745d9c052cb769847591e5f15ffe963f3.png) I’d like to merge here, but the “assignees username” column should merge only unique values, as there’s no point on repeating them. Same for the “custom\_fields type\_config options name”. However, I can’t activate the check to merge only uniques, since then the and “custom\_fields value” lose their correlation to the “custom\_fields name” due to blanks and repeated numbers and I cannot use the values properly. ![image](https://us1.discourse-cdn.com/flex020/uploads/parabola/original/2X/4/47b52bb26463abdb4c1714c7d029626bbb357faf.png) Any idea what can I do? ---
**Author:** ![Axel\_Delgado](https://avatars.discourse-cdn.com/v4/letter/a/ea5d25/32.png) [@Axel\_Delgado](https://community.parabola.io/u/Axel_Delgado)\ **Post date:** [September 1, 2022, 10:04am UTC](https://community.parabola.io/t/expand-json-dataset-into-columns-instead-of-rows/2479/4 "2022-09-01T10:04:09Z")
Bump @Emory_Stainbrook ---
**Author:** ![Ayana\_Usui](https://sea2.discourse-cdn.com/flex020/user_avatar/community.parabola.io/ayana_usui/32/1985_2.png) [@Ayana\_Usui](https://community.parabola.io/u/Ayana_Usui)\ **Post date:** [September 1, 2022, 8:29pm UTC](https://community.parabola.io/t/expand-json-dataset-into-columns-instead-of-rows/2479/5 "2022-09-01T20:29:54Z")
Hey @Axel_Delgado , Sorry to hear that the option to merge only unique values isn’t working on your flow. Would you mind emailing us at [help@parabola.io](mailto:help@parabola.io) a link to your Flow so we can troubleshoot your step? Thanks! Ayana ---
**Author:** ![Axel\_Delgado](https://avatars.discourse-cdn.com/v4/letter/a/ea5d25/32.png) [@Axel\_Delgado](https://community.parabola.io/u/Axel_Delgado)\ **Post date:** [September 2, 2022, 9:13am UTC](https://community.parabola.io/t/expand-json-dataset-into-columns-instead-of-rows/2479/6 "2022-09-02T09:13:18Z")
No worries, I found a solution myself. It’s a bit of a hassle, but it works. I created a sub table by removing all the columns with duplicated values, leaving only ID and the expanded dataset. ![image](https://us1.discourse-cdn.com/flex020/uploads/parabola/original/2X/0/0389e128d742e84233c85bd53cded78fc219684c.png) Then I pivot the columns to get the 229 original rows again, with all the values per ID in a single row. ![image](https://us1.discourse-cdn.com/flex020/uploads/parabola/original/2X/5/580624d624912a74f25371baec23f3196c6c0f6d.png) And then I combine it back with the original table, that I have merged duplicated rows only with Unique values (losing data on those columns, but removing them anyway). It looks like this: ![image](https://us1.discourse-cdn.com/flex020/uploads/parabola/original/2X/1/163cfb034953475c4d94a0746ae59a7951734eea.png)