Update src/streamlit_app.py

src/streamlit_app.py

@@ -1,41 +1,144 @@
+# Libraries
 import streamlit as st
-import os
 import pandas as pd
+import os
+import matplotlib.pyplot as plt
+import seaborn as sns
+from sklearn.model_selection import train_test_split
+from sklearn.ensemble import RandomForestClassifier
+from sklearn.metrics import accuracy_score
 
-st.title("🕵️ File Finder Tool")
+# Title
+st.set_page_config(page_title="Diabetes Prediction", page_icon="🩺")
+st.title("🩺 Diabetes Prediction System")
+st.write("A Machine Learning model trained on the PIMA Indians Diabetes Dataset.")
 
-st.write("Current Working Directory:", os.getcwd())
+# File Loader
+possible_paths = [
+    "diabetes.csv",             # Root folder
+    "src/diabetes.csv",         # src folder
+    "../diabetes.csv",          # Parent folder
+    "/app/diabetes.csv"         # Docker container absolute path
+]
 
-# 1. List ALL files in the entire container (Recursive Search)
-st.subheader("Scanning for 'diabetes.csv'...")
-found = False
-csv_path = ""
+dataset_path = None
 
-# Walk through every folder in the app
-for root, dirs, files in os.walk("."):
-    for file in files:
-        # Check if the file name contains 'diabetes' (case insensitive)
-        if "diabetes" in file.lower():
-            full_path = os.path.join(root, file)
-            st.success(f"✅ FOUND IT! Path: `{full_path}`")
-            found = True
-            csv_path = full_path
+for path in possible_paths:
+    if os.path.exists(path):
+        dataset_path = path
+        break
 
-if not found:
-    st.error("❌ logical_error: The file 'diabetes.csv' is NOT in the container.")
-    st.warning("Make sure you didn't name it 'Diabetes.csv' (Capital D) or 'diabetes.csv.txt'.")
-    
-    st.subheader("Listing ALL files I can see:")
-    # List all files so you can see what IS there
-    for root, dirs, files in os.walk("."):
-        for file in files:
-            st.code(os.path.join(root, file))
+if dataset_path:
+    pass
 else:
-    # 2. Try to load it using the path we found
-    st.subheader("Trying to load data...")
-    try:
-        df = pd.read_csv(csv_path)
-        st.write("Preview of data:", df.head())
-        st.success("🎉 It works! Now you know the correct path to use.")
-    except Exception as e:
-        st.error(f"Found the file but failed to read it: {e}")
+    st.warning("'diabetes.csv' not found. Using dummy data for demonstration.")
+    dummy_data = {
+        'Pregnancies': [6, 1, 8, 1, 0, 5, 3, 10],
+        'Glucose': [148, 85, 183, 89, 137, 116, 78, 115],
+        'BloodPressure': [72, 66, 64, 66, 40, 74, 50, 0],
+        'SkinThickness': [35, 29, 0, 23, 35, 0, 32, 0],
+        'Insulin': [0, 0, 0, 94, 168, 0, 88, 0],
+        'BMI': [33.6, 26.6, 23.3, 28.1, 43.1, 25.6, 31.0, 35.3],
+        'DiabetesPedigreeFunction': [0.627, 0.351, 0.672, 0.167, 2.288, 0.201, 0.248, 0.134],
+        'Age': [50, 31, 32, 21, 33, 30, 26, 29],
+        'Outcome': [1, 0, 1, 0, 1, 0, 1, 0]
+    }
+    df_dummy = pd.DataFrame(dummy_data)
+    df_dummy.to_csv("diabetes.csv", index=False)
+    dataset_path = "diabetes.csv"
+
+# Train Model
+@st.cache_resource
+def train_model(path):
+    df = pd.read_csv(path)
+    
+    # Data Split
+    X = df.drop(columns=['Outcome'])
+    y = df['Outcome']
+    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)
+
+    # Train Random Forest
+    model = RandomForestClassifier(random_state=42)
+    model.fit(X_train, y_train)
+    
+    # Calculate accuracy
+    acc = accuracy_score(y_test, model.predict(X_test))
+    
+    return model, acc, df  # Return df as well for plotting
+
+# Load and Train
+model, accuracy, df = train_model(dataset_path)
+
+# Display Accuracy
+st.write(f"**Model Accuracy:** {accuracy * 100:.2f}%")
+
+# UI/UX
+if st.checkbox("Show Model Insights & Graphs"):
+    st.subheader("Data Visualizations")
+    
+    col1, col2 = st.columns(2)
+    
+    with col1:
+        st.write("**Outcome Distribution**")
+        fig1, ax1 = plt.subplots()
+        df['Outcome'].value_counts().plot.pie(autopct='%1.1f%%', ax=ax1, labels=['Non-Diabetic', 'Diabetic'], colors=['#66b3ff', '#ff9999'])
+        st.pyplot(fig1)
+
+    with col2:
+        st.write("**Correlation Heatmap**")
+        fig2, ax2 = plt.subplots()
+        sns.heatmap(df.corr(), annot=False, cmap='coolwarm', ax=ax2)
+        st.pyplot(fig2)
+        
+    st.write("**Feature Importance (What matters most?)**")
+    fig3, ax3 = plt.subplots()
+    importances = pd.Series(model.feature_importances_, index=df.drop(columns=['Outcome']).columns)
+    importances.sort_values().plot(kind='barh', color='teal', ax=ax3)
+    st.pyplot(fig3)
+
+st.markdown("---")
+
+# Inputs
+st.sidebar.header("Patient Data")
+
+def user_input_features():
+    pregnancies = st.sidebar.slider('Pregnancies', 0, 17, 3)
+    glucose = st.sidebar.slider('Glucose', 0, 199, 117)
+    bp = st.sidebar.slider('Blood Pressure', 0, 122, 72)
+    skin = st.sidebar.slider('Skin Thickness', 0, 99, 23)
+    insulin = st.sidebar.slider('Insulin', 0, 846, 30)
+    bmi = st.sidebar.slider('BMI', 0.0, 67.1, 32.0)
+    dpf = st.sidebar.slider('Diabetes Pedigree Function', 0.078, 2.42, 0.3725)
+    age = st.sidebar.slider('Age', 21, 81, 29)
+    
+    data = {
+        'Pregnancies': pregnancies,
+        'Glucose': glucose,
+        'BloodPressure': bp,
+        'SkinThickness': skin,
+        'Insulin': insulin,
+        'BMI': bmi,
+        'DiabetesPedigreeFunction': dpf,
+        'Age': age
+    }
+    features = pd.DataFrame(data, index=[0])
+    return features
+
+input_df = user_input_features()
+
+st.subheader('User Input parameters')
+st.write(input_df)
+
+# Prediction
+if st.button('Predict Diabetes Risk'):
+    prediction = model.predict(input_df)
+    prediction_proba = model.predict_proba(input_df)
+
+    st.subheader('Prediction Result')
+    
+    if prediction[0] == 1:
+        st.error(f' Positive (High Risk of Diabetes)')
+        st.write(f"Confidence: {prediction_proba[0][1] * 100:.2f}%")
+    else:
+        st.success(f' Negative (Low Risk of Diabetes)')
+        st.write(f"Confidence: {prediction_proba[0][0] * 100:.2f}%")