dtale not working in DataBricks (Azure)

[OhStan]

Hi,

Need help running on Azure DataBricks. If I try it with following, it does not return the table but the following. Without "ignore_duplicate=True", a link will be provided, but will lead to a page that is not reachable. How do we get the table display?

Input: import dtale #SPARK_KOALAS_AUTOPATCH=false d=dtale.show(df_py, ignore_duplicate=True) #df_py is a pandas dataframe

Output: 2020-09-01 00:43:06,875 - INFO - Received command c on object id p0 2020-09-01 00:43:06,875 - INFO - Received command c on object id p0 2020-09-01 00:43:06,875 - INFO - Received command c on object id p0 2020-09-01 00:43:06,876 - INFO - Received command c on object id p0 2020-09-01 00:43:06,876 - INFO - Received command c on object id p0 2020-09-01 00:43:06,877 - INFO - Received command c on object id p0 2020-09-01 00:43:06,877 - INFO - Received command c on object id p0 2020-09-01 00:43:06,877 - INFO - Received command c on object id p0 2020-09-01 00:43:06,877 - INFO - Received command c on object id p0

Thanks.

[aschonfeld]

So unfortunately, it looks like data bricks doesn't have great support for running flask processes within their notebooks (see this post).

Unfortunately, D-Tale spins up a flask process when it is instantiated and data bricks doesn't have great support for separate flask processes. Here's another post detailing that data bricks is mainly for running calculations and not hosting web applications.

[tkhan3]

can we have more information about running dtale on databricks ?

[aschonfeld]

@tkhan3 unfortunately the Databricks team did not return my requests for help on this issue and my free databricks account ran out. What's even better is that they've deleted the previous posts I referenced on the issue.

You can try reaching out to them directly since you might have support bundled into your databricks account. If they do respond and need more information on how D-Tale works feel free to reach out to me