Using Python UDF in Delta live table

[Sabtout]

Hello,

I tried running a python UDF in a Delta Live Table workflow in Advanced mode but it did not run and gave the "Python UDF is not supported in your environment" error.

Can I get a clear picture if the Python External UDFs are supported or not?

[Kaniz]

Hi @Sabtout , you can use Python user-defined functions (UDFs) in your SQL queries, but these UDFs must be defined in Python files in the same pipeline before calling them in SQL source files. If you're encountering an error stating "Python UDF is not supported in your environment", it may be due to an environment or configuration issue, or a limitation in the current Delta Live Tables setup.

[Sabtout]

Hi @Kaniz 

I ran this SQL query in my Catalog (I'm using Unity Catalog) :

CREATE OR REPLACE FUNCTION cat_projint_dev.silver.GetEditor(prompt STRING)

RETURNS STRING

LANGUAGE PYTHON

AS $$

print(prompt)

$$

Then I ran a Delta Live Table workflow using Unity Catalog plugged in my silver catalog like this :

@dlt.table(name="editors_bronze", temporary = True)

def editors_bronze():

    return spark.sql("SELECT prompt,cat_projint_dev.bronze.GetEditor(prompt) AS EDITOR_NAME FROM LIVE.editors_temp_bronze")

And then I'm having this issue :

[FEATURE_UNAVAILABLE] Python UDF is not supported in your environment. To use this feature, please contact Databricks Support.

Thank you for your help ! 🙂