You are viewing a plain text version of this content. The canonical link for it is here.
Posted to dev@superset.apache.org by vimal goyal <vi...@gmail.com> on 2018/02/22 16:28:40 UTC

Integrating Jupyter data model with Superset

Hi,

First of all thanks to everyone who is contributing on Apache Superset, a
really beautiful data visualization tool with the rich features.

Currently, I am working on a project where I have a situation where I need
some guidance.

I have three systems, one has the database, another has jupyter notebook
and third has superset.

First, we create the data model in R/Python using jupyter nb. Now we would
like to run these data model on the database via superset so that we can
visualize the data. And each time when we run the data model on data it
should visualize the latest data from the database.

We know that;

1. We can directly connect superset with database and play with data.
2. or can run data model manually at the backend and generate the output
data as csv and can visualize the result on the superset.

We can not go with the 1st approach because we have the separate users for
both Jupyter and superset. Jupiter user cannot access superset to create
the data model, slices, and dashboard. Also, we do not want to expose raw
data with superset user and superset user should able to only see the
graphs/charts generated from the output of the data models.

We can also not go with 2nd approach because every time when we want to see
the latest data, we have to first run data model manually.

So is there any solution to my problem where I can run data models in the
background whenever I reload the dashboard and I can see refreshed data.

Any lead/suggestion will be highly appreciated.

Thanks,
Vimal