# How to get all the non-empty coordinates from sparse array? 

**URL:** https://forum.tiledb.com/t/how-to-get-all-the-non-empty-coordinates-from-sparse-array/284
**Category:** Uncategorized
**Created:** [December 10, 2020, 3:15pm UTC](https://forum.tiledb.com/t/how-to-get-all-the-non-empty-coordinates-from-sparse-array/284 "2020-12-10T15:15:02Z")
**Posts on this page:** 2
**Page:** 1

<div class="post-metadata">

### Author: ![WarBean](https://yyz1.discourse-cdn.com/flex035/user_avatar/forum.tiledb.com/warbean/32/119_2.png) [@WarBean](https://forum.tiledb.com/u/WarBean)
#### Post date: [December 10, 2020, 3:15pm UTC](https://forum.tiledb.com/t/how-to-get-all-the-non-empty-coordinates-from-sparse-array/284/1 "2020-12-10T15:15:03Z")

</div>

Suppose I create a sparse array:

```auto
import tiledb
import numpy as np

path = 'test.tiledb'
dim1 = tiledb.Dim(name='dim1', domain=(None, None), tile=None, dtype=np.bytes_)
dim2 = tiledb.Dim(name='dim2', domain=(None, None), tile=None, dtype=np.bytes_)
attr = tiledb.Attr(name='attr', dtype=np.float32)
dom = tiledb.Domain(dim1, dim2)
schema = tiledb.ArraySchema(domain=dom, sparse=True, attrs=[attr])
tiledb.Array.create(path, schema)

with tiledb.open(path, 'w') as array:
    array['a1', 'b1'] = 1
    array['a1', 'b2'] = 2
    array['a2', 'b1'] = 3
    array['a3', 'b1'] = 4

```

How can I retrieve all the “keys” from `array` , like `array.dim('dim1').values()` or `array.coordinates('dim1')` , which should return `['a1', 'a2', 'a3']` ?

---

<div class="post-metadata">

### Author: ![ihnorton](https://yyz1.discourse-cdn.com/flex035/user_avatar/forum.tiledb.com/ihnorton/32/23_2.png) [@ihnorton](https://forum.tiledb.com/u/ihnorton)
#### Post date: [December 10, 2020, 3:16pm UTC](https://forum.tiledb.com/t/how-to-get-all-the-non-empty-coordinates-from-sparse-array/284/2 "2020-12-10T15:16:54Z")

</div>

Hi @WarBean,

This message got stuck in the review queue - sorry about that! I will cross-link the GitHub issue where we are discussing and continue there:

> <https://github.com/TileDB-Inc/TileDB-Py/issues/441>
>
> Suppose I create a sparse array:
> import tiledb
> import numpy as np
> 
> path = 'test.tiledb'
> dim1 = tiledb.Dim(name='dim1', domain=(None, None), tile=None, dtype=np.bytes\_)
> dim2 = tiledb.Dim(name='dim2', domain=(None,...

Best,  
Isaiah