Pycaret plot_model not displaying

Using Streamlit
1

Summary

I’m trying to display the feature importance chart from the pycaret library, but the chart is not displayed

Steps to reproduce

Code snippet:

import streamlit as st
from pycaret.regression import *
import pandas as pd
import os

if os.path.exists('./dataset.csv'):
    df = pd.read_csv('dataset.csv', index_col=None)
    chosen_target = st.selectbox('Choose the Target Column', df.columns)
    if st.button('Run Modelling'):
        s = setup(df, target=chosen_target, session_id = 123)
        best_model = compare_models()
        plot_model(best_model, plot='feature', display_format='streamlit')

Apart from the graph, the code works and the model is created correctly. Sample data: Vehicle dataset | Kaggle

Expected behavior:

Need to see feature importance plot like in PyCaret inside Streamlit - #16 by ferdy

Actual behavior:

The plot is not displayed.

Debug info

  • Streamlit version: 1.27.2
  • Python version: 3.9.13
  • Pycaret 3.10
2

Hey @Keszzz! :wave:

It’s not something I’ve tried, but instead of directly plotting with PyCaret’s plot_model function, you might consider using Streamlit’s st.pyplot function to display it.

Also, have you checked that you’re using the latest versions of both Streamlit and PyCaret?

Best,
Charly

3

Here’s one way to get it to work (not sure why display_format=“streamlit” isn’t working)

        img = plot_model(
            best_model, plot="feature", display_format="streamlit", save=True
        )
        st.image(img)
4

Great walkaround. Works perfectly.