# User modified data table saveing

**URL:** <https://community.plotly.com/t/user-modified-data-table-saveing/38770>\
**Category:** Dash Python\
**Created:** [May 2, 2020, 8:42am UTC](https://community.plotly.com/t/user-modified-data-table-saveing/38770 "2020-05-02T08:42:19Z")\
**Posts on this page:** 3\
**Page:** 1

<div class="post-metadata">

**Author:** ![Garik](https://avatars.discourse-cdn.com/v4/letter/g/7ba0ec/32.png) [@Garik](https://community.plotly.com/u/Garik)\
**Post date:** [May 2, 2020, 8:42am UTC](https://community.plotly.com/t/user-modified-data-table-saveing/38770/1 "2020-05-02T08:42:19Z")

</div>

how to save user modified data table?

---

<div class="post-metadata">

**Author:** ![psip](https://avatars.discourse-cdn.com/v4/letter/p/8edcca/32.png) [@psip](https://community.plotly.com/u/psip)\
**Post date:** [May 2, 2020, 11:44am UTC](https://community.plotly.com/t/user-modified-data-table-saveing/38770/2 "2020-05-02T11:44:26Z")

</div>

If you just want to load the table as an .csv or .xlsx, set export\_format to ‘csv’ or ‘xlsx’, accordingly (see [https://dash.plotly.com/datatable/reference](https://dash.plotly.com/datatable/reference)).

If you want to persist the state of the table locally, use persistence=True with ‘data’ in persisted\_props.

If you want to more permanently store table data, I recommend using [Postgresql](https://www.postgresql.org) with [psycopg2](https://www.psycopg.org/docs/) or [SQLAlchemy](https://www.sqlalchemy.org). Probably most straightforward is to store table data as json (for example in a callback when the user clicks a button).

---

<div class="post-metadata">

**Author:** ![Garik](https://avatars.discourse-cdn.com/v4/letter/g/7ba0ec/32.png) [@Garik](https://community.plotly.com/u/Garik)\
**Post date:** [May 2, 2020, 6:06pm UTC](https://community.plotly.com/t/user-modified-data-table-saveing/38770/3 "2020-05-02T18:06:34Z")

</div>

Here is the code.

```auto
dash_table.DataTable(			      
                   id='kpi_retail_monthly',
                   data=df.to_dict('rows'),
		   persistence=True,
                   rsisted_props =['data'],
		   columns=[.......

```