Python Streamlit Error: Solving Issues

I’m trying to create a Streamlit page in Google Colab to deploy a simple spam-ham classifier, but I keep running into an error, and I can’t figure out what’s wrong. I’ve installed Streamlit and pyngrok, set up the necessary imports, and written the main app code for the Streamlit interface. My code loads a pre-trained model with joblib, takes user input, and predicts whether the message is spam or ham when I hit the “Predict” button. However, when I try to run the last block to start the Streamlit app and open it with ngrok, it throws an error. I’ve checked the code multiple times, but I can’t seem to pinpoint the issue.

Error Code:

code!pip install streamlit --quiet 
!pip install pyngrok==4.1.1 
from pyngrok import ngrok

%%writefile app.py 
import streamlit as st
import joblib 
st.title('Spam Ham Deployment') 
test_model=joblib.load('spam_ham') 
ip=st.text_input('Enter your message')
 op=test_model.predict([ip]) 
if st.button('Predict')   
    st.title(op)

!nohup streamlit run app.py & 
url=ngrok.connect(port='8501')
 url

The code you’ve shared is intended to deploy a machine learning model for classifying messages as “spam” or “ham” using Streamlit and Pyngrok. The goal is to create a simple web app where users can input a message and receive a classification (spam or ham). However, the code contains several syntax and indentation errors that need correction to run properly.

Issues and Solutions:

1. Indentation:
   • The line op=test_model.predict([ip]) is not aligned properly, which causes a syntax error.
   • url should be aligned correctly to ensure the code flows without indentation errors.
2. Missing Colon:
   • The if statement in if st.button('Predict') lacks a colon, which is essential for Python syntax.
3. Streamlit Button Action:
   • The prediction should be executed only when the “Predict” button is clicked.
4. Running Streamlit with Ngrok:
   • Ngrok is used to create a public URL for the Streamlit app, making it accessible over the internet.

Correct Code:

code# Install necessary packages
!pip install streamlit --quiet
!pip install pyngrok==4.1.1

from pyngrok import ngrok

# Write the Streamlit app to a Python file
%%writefile app.py
import streamlit as st
import joblib 

# Set the title of the app
st.title('Spam Ham Deployment') 

# Load the trained model
test_model = joblib.load('spam_ham') 

# Get user input
ip = st.text_input('Enter your message')

# Predict and display result when button is clicked
if st.button('Predict'):
    op = test_model.predict([ip])  # Predict spam or ham
    st.title(op[0])  # Display the prediction

# Run the Streamlit app with Ngrok
!nohup streamlit run app.py & 
url = ngrok.connect(port='8501')
url

Explanation:

1. Package Installation:
   • streamlit and pyngrok are installed using pip. streamlit is the web app framework, while pyngrok is used to make the app accessible from a public URL.
2. Ngrok Setup:
   • After importing ngrok, the code writes the Streamlit app code to a file named app.py.
3. Streamlit App Setup:
   • The Streamlit app begins by importing necessary libraries: streamlit for the app interface and joblib to load the saved model.
   • The app title is set to “Spam Ham Deployment”.
4. Model Loading:
   • The joblib.load('spam_ham') command loads a pre-trained model (expected to be saved as spam_ham), which will classify the input message.
5. User Input and Prediction:
   • st.text_input creates a text box for the user to input a message.
   • if st.button('Predict'): waits for the user to click the “Predict” button. When clicked, it runs the model’s predict function on the input and displays the prediction (spam or ham).
6. Running Streamlit and Exposing with Ngrok:
   • The code uses !nohup streamlit run app.py & to run the Streamlit app in the background.
   • ngrok.connect(port='8501') opens a public URL for the app at port 8501, enabling external access to the Streamlit app.

Key Points:

• Indentation and colon syntax are essential for Python code structure.
• Streamlit and Ngrok combination allows for creating and sharing web applications without a full web server setup.

By using this corrected code, you should be able to run your Streamlit app and access it publicly through the Ngrok URL.