Do you know the History of Data Science? READ MORE

Overview of all the bundles

kim13
Level 1
Overview of all the bundles

Hi All, 

Do you know how to get an overview of all the bundles available? 

What if i have 1000 bundles and I need an overview of all the bundles ? 

Do you think it is possible to extract all the bundles info and have it in a table inside dataiku to have an overview?

Thank you. 

Best Regards, 

Orwin

0 Kudos
3 Replies
HarizoR
Dataiker
Dataiker

Hi,

You can programmatically list the exported bundles of a given project using the following code:

 

import dataiku
client = dataiku.api_client()
project = client.get_project("TEST")
bundles = project.list_exported_bundles()

 

The output of this code (in the bundles variable) will be a list of dictionaries, each of those containing the detailed description of a distinct bundle. From there, you will be able to apply any filtering operation you need, or even write it into a Dataset.

Best,

Harizo

0 Kudos
kim13
Level 1
Author

Thank you HarizoR.

 

Do you have any idea why Dataiku did not include the overview of bundles in their rollouts ? 

I find it very strange that the open source tools like DVC powered by GIT also has this overview of their versions but a closed source like Dataiku does not. Please share if you have any idea why Dataiku does not have this overview in their default features. 

Thanks!

0 Kudos
HarizoR
Dataiker
Dataiker

Hi kim13,

As stated in my previous message, the overview of bundles is de facto available in Dataiku DSS via the API. Instead of embedding the access to this information in a UI, the ability to leverage the API allows you much more flexibility on how you want to process that information:

- you can combine it with the Datasets API in order to write that information in a Dataset directly accessible from the Flow using any connection you have available,

- you can create a custom UI thanks to webapps if you are looking for something more interactive,

- the fact that the API is RESTful even allows you to send the information to any 3rd-party tool that can send HTTP requests.

On a final note, we are actively maintaining the Python client that sits on top of the API, the client itself is open-source and available in this GitHub repository.

 

Best,

Harizo.

 

0 Kudos
A banner prompting to get Dataiku DSS