app.py

"""
Iris Classification Streamlit App

This module loads the pre-trained Iris classifier pipeline from disk,
defines iris class mappings, and sets up a Streamlit interface that allows
users to input iris flower features and obtain predictions for the iris species.
"""

import streamlit as st
import pickle
import numpy as np

# Load the saved pipeline (generated by train_model.py)
try:
    with open("iris_pipeline.pkl", "rb") as f:
        pipeline = pickle.load(f)
except Exception as e:
    st.error(f"Error loading the model pipeline: {e}")
    st.stop()

# Mapping of iris class indices to human-readable species names
iris_classes = {0: "setosa", 1: "versicolor", 2: "virginica"}

# Application title and description
st.title("Iris Flower Classification")
st.write("Enter the iris features to predict the species:")

# Input fields for iris features
sepal_length = st.number_input("Sepal Length", min_value=0.0, max_value=10.0, value=5.1)
sepal_width  = st.number_input("Sepal Width", min_value=0.0, max_value=10.0, value=3.5)
petal_length = st.number_input("Petal Length", min_value=0.0, max_value=10.0, value=1.4)
petal_width  = st.number_input("Petal Width", min_value=0.0, max_value=10.0, value=0.2)

# When the Predict button is clicked, perform the prediction
if st.button("Predict"):
    # Prepare the feature vector as a 2D array
    features = np.array([[sepal_length, sepal_width, petal_length, petal_width]])
    try:
        # Perform prediction using the loaded pipeline
        prediction = pipeline.predict(features)
        species = iris_classes[prediction[0]]
        st.write("Predicted Iris Species:", species)
    except Exception as e:
        st.error(f"Error during prediction: {e}")