# Help in restructuring table

**URL:** <https://community.influxdata.com/t/help-in-restructuring-table/22370>\
**Category:** Fluxlang\
**Created:** [November 2, 2021, 10:56am UTC](https://community.influxdata.com/t/help-in-restructuring-table/22370 "2021-11-02T10:56:04Z")\
**Posts on this page:** 2\
**Page:** 1

<div class="post-metadata">

**Author:** ![sauvant](https://sea1.discourse-cdn.com/flex023/user_avatar/community.influxdata.com/sauvant/32/3703_2.png) [@sauvant](https://community.influxdata.com/u/sauvant)\
**Post date:** [November 2, 2021, 10:56am UTC](https://community.influxdata.com/t/help-in-restructuring-table/22370/1 "2021-11-02T10:56:04Z")

</div>

Hi there,

having a quite hard time to get used to flux coming from SQL 🤪  
Could you please help me with a table restructuring task?

This is example data for my input table format:

id val0 val1 val2  
100 1 5 7  
101 8 5 4  
102 9 9 1

And this is what I want to get:

id valno val  
100 0 1  
100 1 5  
100 2 7  
101 0 8  
101 1 5  
101 2 4  
102 0 9  
102 1 9  
102 2 1

Little bit like a “reverse pivot” is what I need…  
Any help is appreciated!

Best,  
Keith

---

<div class="post-metadata">

**Author:** ![scott](https://sea1.discourse-cdn.com/flex023/user_avatar/community.influxdata.com/scott/32/16492_2.png) [@scott](https://community.influxdata.com/u/scott)\
**Post date:** [November 29, 2021, 9:28pm UTC](https://community.influxdata.com/t/help-in-restructuring-table/22370/2 "2021-11-29T21:28:17Z")

</div>

@sauvant There is an issue about this exact functionality. Please feel free to comment on it: [Add unpivot functionality · Issue #2539 · influxdata/flux · GitHub](https://github.com/influxdata/flux/issues/2539)

It is possible to do what you want, but it isn’t very straight forward. The example below uses `array.from` to build a stream of tables that matches your data, so you don’t actually need it when you do this operation on your actual data, but it creates 3 separate streams of tables (one for each value number), restructures them to what you need, unions them back together and applies the necessary sorting:

```plaintext
import "array"

data = array.from(rows: [
    {id: 100, val0: 1, val1: 5, val2: 7},
    {id: 101, val0: 8, val1: 5, val2: 4},
    {id: 102, val0: 9, val1: 9, val2: 1},
])

d0 = data |> keep(columns: ["id", "val0"]) |> map(fn: (r) => ({id: r.id, valno: 0, val: r.val0}))
d1 = data |> keep(columns: ["id", "val1"]) |> map(fn: (r) => ({id: r.id, valno: 1, val: r.val1}))
d2 = data |> keep(columns: ["id", "val2"]) |> map(fn: (r) => ({id: r.id, valno: 2, val: r.val2}))

union(tables: [d0, d1, d2])
  |> sort(columns: ["id", "valno"])

```