# optimize ungroup columns operation in kdb

**URL:** <https://forum.kx.com/t/optimize-ungroup-columns-operation-in-kdb/10702>\
**Category:** Community Support\
**Tags:** kdb-and-q\
**Created:** [January 1, 2016, 8:44pm UTC](https://forum.kx.com/t/optimize-ungroup-columns-operation-in-kdb/10702 "2016-01-01T20:44:00Z")\
**Posts on this page:** 1\
**Page:** 1

<div class="post-metadata">

**Author:** ![Sam11](https://avatars.discourse-cdn.com/v4/letter/s/c6cbf5/32.png) [@Sam11](https://forum.kx.com/u/Sam11)\
**Post date:** [January 1, 2016, 8:44pm UTC](https://forum.kx.com/t/optimize-ungroup-columns-operation-in-kdb/10702/1 "2016-01-01T20:44:00Z")

</div>

I am trying to flatten some columns of a table into a single column. My current approach is to enlist those columns into a single column and then use ungroup on table to flatten it. But this approach is very inefficient, takes more than 2 seconds on 1300000 rows and 3 columns in this example. My actual table has more rows and about 20 columns to flatten. Is there a way to optimize this or an alternative approach?

```
/ flatten v1,v2 and v3 columns into one column va:([]cust:n?`8; v1:n?100; v2:n?100; v3:n?100)\t ungroup ?[a;();0b;`cust`v!(`cust;(flip; plist,`v1`v2`v3))]2079

```