Discover the winners & finalists of the 2022 Dataiku Frontrunner Awards!READ THEIR USE CASES

Create and populate Editable Datasource with API

benmoss
Level 2
Create and populate Editable Datasource with API

Hi!

I'm looking at the different API end-points in regards to creating datasources.

Specifically in this case I am looking for a method of initially creating an editable datasource and then populating that editable datasource with data.

The first part seems possible but I can't see any documentation to say that the 2nd part is also. Has anyone tried to do anything similar?

Simple table to load as an example.

PersonAge
Ben31
Charlotte28

 

Ben

0 Kudos
1 Reply
Emma
Dataiker

Hey @benmoss ,

Great questions. Editable Datasets are really meant to be modified through the UI. 

However, if you have need to use them then a workaround was detailed in this Community post: https://community.dataiku.com/t5/Using-Dataiku/Can-we-write-to-a-editable-dataset-using-Python-API/m... 

If your use case can be served without having an Editable Dataset (you are hoping to create and then populate a dataset entirely programmatically) you can do so with the generic Dataiku Dataset API: 

# Get the Dataiku client 
client = dataiku.api_client()
project = client.get_project("ProjectKey")

# Create the Dataset
builder = project.new_managed_dataset("mydatasetname")
builder.with_store_into("TargetLocation")
dataset = builder.create()

# Now modify the Dataset in some way
mydatasetname = ... 

Hope that helps, 

Emma 

0 Kudos