Survey banner
The Dataiku Community is moving to a new home! We are temporary in read only mode: LEARN MORE

import dataiku Error Message

mbanate
Level 2
import dataiku Error Message

Here is the error message I get when I am trying to import dataiku.

errorMessage.png

My installed packages are shown in the following picture. In fact for creating a kernel I have used Python 3.9, Pandas 1.3, in addition to the latest versions of matplotlib, numpy, sklearn, xgboost, hyperopt, kds, sklearn2pmml.

installedPackages.png

 

 

 

0 Kudos
5 Replies
CatalinaS
Dataiker
​Hi @mbanate,
 
The error you are receiving comes from deprecated 'np.bool' being entirely removed as of numpy 1.24.0. A fix for this new numpy version was introduced in DSS 11.2.1.

Another solution is to specify an older numpy version on your code env:
 1) Go to Administration > Code Envs > Select your code env
 2) Add the following in 'Packages to install':  numpy < 1.24
 3) Click 'UPDATE'
0 Kudos
mbanate
Level 2
Author

Thanks for your reply. I updated the numpy and now I get this error.

newError.png

0 Kudos
CatalinaS
Dataiker

The new error is about urllib3 package which is included in the base packages of the code envs. The problem comes from the latest urllib3 package release, which dropped support for OpenSSL < 1.1.1. The fix for this issue has been delivered in DSS 11.4.3.

The workaround is to add the following constraints for the urllib3 into the 'packages to install' list of the code env used like below and rebuild the code env:

urllib3<2.0,>=1.21.1

Upgrading DSS would fix both of the issues.

0 Kudos
mbanate
Level 2
Author

Thanks @CatalinaS 

I was using the default DDS, how can I upgrade that? Doesn't it automatically use the latest version?

0 Kudos

@mbanate wrote:

I was using the default DDS, how can I upgrade that? Doesn't it automatically use the latest version?


It depends on which Dataiku edition you have:

https://www.dataiku.com/product/plans-and-features/

The Windows / MacOS versions update when you launch them. The Cloud versions are updated by Dataiku. Enterprise on-prem it's up to you to update.

0 Kudos