Statsmodel OLS.fit generates "Error displaying widget: undefined"

[JU1]

Hi,

I am running some code which fits a linear model via Statsmodel. Everytime I run the .fit function it generates the error: "Error displaying widget: undefined".

I can reproduce the error via the simple code below.

I am not sure what is causing this issue and I cannot find any similar issues online. Is anybody aware of the issue causing this error?

Node type: Standard_DS3_v2

Databricks runtime version: 15.1 ML

Statsmodel version: 0.14.0

import statsmodels.api as sm

import pandas as pd

# Create sample data

data = {

    "X": [1, 2, 3, 4, 5],

    "Y": [2, 4, 5, 4, 5]

}

df = pd.DataFrame(data)

# Define independent and dependent variables

X = sm.add_constant(df["X"])  # Add constant for intercept

y = df["Y"]

# Fit OLS model

model = sm.OLS(y, X).fit()

[Alberto_Umana · Accepted solution]

Hi @JU1,

The issue is likely related to an issue with the ipywidgets library, which is used for rendering interactive widgets in Jupyter notebooks. This issue has been observed in compatibility problems between different versions of ipywidgets and other libraries.

Can you run this command and send the output?

import ipywidgets as widgets
print(widgets.__version__)

Also can you try on another DBR version like 14.3

[JU1 · Accepted solution]

Hi,

It was indeed the DBR which caused the issue. I changed it to 16.1 and the widgets now load.

Thanks

[Alberto_Umana]

Awesome, which DBR were you using before?