Update app.py

app.py

@@ -1,11 +1,3 @@
-# -*- coding: utf-8 -*-
-"""
-Created on Wed Apr 23 22:57:11 2025
-
-@author: wIN 10
-"""
-
-# app.py
 import streamlit as st
 import pandas as pd
 import numpy as np
@@ -13,35 +5,37 @@ from datetime import date
 from sklearn.ensemble import RandomForestClassifier
 from catboost import CatBoostClassifier
 from sklearn.preprocessing import MinMaxScaler
-from sklearn.metrics import classification_report, accuracy_score, precision_score, recall_score, f1_score, roc_auc_score, confusion_matrix
+from sklearn.metrics import classification_report
 from imblearn.under_sampling import RandomUnderSampler
 import matplotlib.pyplot as plt
 import os
 
-# Load or initialize dataset
+# Load dataset (use relative path)
 DATA_PATH = "updated_dataset.csv"
 if os.path.exists(DATA_PATH):
     df = pd.read_csv(DATA_PATH)
 else:
-    st.error("Dataset not found. Please upload 'updated_dataset.csv'.")
+    st.error("❌ Dataset 'updated_dataset.csv' not found. Please upload it.")
 
 features = ['Temperature (°C)', 'Humidity (%)', 'Air Pressure (kPa)', 'Wind Speed (m/s)', 'Wind Direction (°)']
 target = 'Rain_Label'
 df = df[~(df[features] == -999).any(axis=1)]
+
+# Normalize
 scaler = MinMaxScaler()
 df[features] = scaler.fit_transform(df[features])
 
 X = df[features]
 y = df[target]
-rus = RandomUnderSampler(random_state=42)
-X_bal, y_bal = rus.fit_resample(X, y)
+X_bal, y_bal = RandomUnderSampler(random_state=42).fit_resample(X, y)
 
+# Train models
 rf = RandomForestClassifier(n_estimators=100, random_state=42).fit(X_bal, y_bal)
 cat = CatBoostClassifier(verbose=0, random_state=42).fit(X_bal, y_bal)
 
-# --- Streamlit UI ---
+# Streamlit UI
 st.title("🌧 Rainfall Early Warning System")
-st.markdown("Use weather inputs to predict rainfall & submit real-time data.")
+st.markdown("Predict rainfall using daily weather inputs. Submit values below to simulate forecasting.")
 
 st.sidebar.header("Weather Input")
 temp = st.sidebar.slider("Temperature (°C)", 10.0, 50.0, 30.0)
@@ -52,7 +46,6 @@ wind_dir = st.sidebar.slider("Wind Direction (°)", 0, 360, 180)
 rain_label = st.sidebar.selectbox("Did it Rain Today?", ["No", "Yes"])
 rain_val = 1 if rain_label == "Yes" else 0
 
-# Input Data
 input_row = {
     'Temperature (°C)': temp,
     'Humidity (%)': humidity,
@@ -60,42 +53,31 @@ input_row = {
     'Wind Speed (m/s)': wind_speed,
     'Wind Direction (°)': wind_dir,
     'Rainfall (mm)': 0.0,
-    'Rain_Label': rain_val
+    'Rain_Label': rain_val,
+    'Date': str(date.today())
 }
 
-scaled_input = scaler.transform(pd.DataFrame([input_row])[features])
-
 # Predict
+scaled_input = scaler.transform(pd.DataFrame([input_row])[features])
 rf_pred = rf.predict(scaled_input)[0]
 cat_pred = cat.predict(scaled_input)[0]
 rf_prob = rf.predict_proba(scaled_input)[0][1]
 cat_prob = cat.predict_proba(scaled_input)[0][1]
 
-# Show result
-st.subheader("🔍 Prediction")
+st.subheader("🔍 Prediction Results")
 st.markdown(f"**Random Forest:** {'🌧 Rain' if rf_pred == 1 else '☀️ No Rain'} (Prob: {rf_prob:.2f})")
 st.markdown(f"**CatBoost:** {'🌧 Rain' if cat_pred == 1 else '☀️ No Rain'} (Prob: {cat_prob:.2f})")
 
-# Warning label
-def warning(prob):
+def warning_label(prob):
     if prob < 0.1: return "✅ No Rain"
     elif prob < 0.5: return "🌦 Light Rain"
     elif prob < 0.8: return "🌧 Moderate Rain"
     else: return "🌧🌧 Heavy Rain"
 
-st.success(f"📢 Early Warning: {warning((rf_prob + cat_prob) / 2)}")
-
-# --- Real-Time Submission ---
-st.subheader("📥 Submit Today’s Weather Data")
-if st.button("➕ Add to Dataset"):
-    input_row["Date"] = str(date.today())
-    df_new = pd.read_csv(DATA_PATH)
-    df_new = pd.concat([df_new, pd.DataFrame([input_row])], ignore_index=True)
-    df_new.to_csv(DATA_PATH, index=False)
-    st.success("✅ Today's data added successfully!")
+st.success(f"📢 Early Warning: {warning_label((rf_prob + cat_prob) / 2)}")
 
-# --- Model Evaluation ---
-st.subheader("📊 Evaluation Metrics")
+# (Optional) Show evaluation
+st.subheader("📊 Model Evaluation")
 rf_preds = rf.predict(X)
 cat_preds = cat.predict(X)
 st.markdown("**Random Forest:**")