Create app.py

app.py

@@ -0,0 +1,161 @@
+import streamlit as st
+import pandas as pd
+import random
+import joblib
+from sklearn.ensemble import RandomForestClassifier
+from sklearn.preprocessing import StandardScaler
+from sklearn.model_selection import train_test_split
+from sklearn.metrics import classification_report
+import matplotlib.pyplot as plt
+import seaborn as sns
+import numpy as np
+import os
+
+# Load cattle data
+DATA_FILE = 'cattle_data.csv'
+
+def load_data():
+    # Check if the dataset exists
+    if os.path.exists(DATA_FILE):
+        return pd.read_csv(DATA_FILE)
+    else:
+        return pd.DataFrame(columns=['TAG ID', 'BREED', 'AGE', 'SURGERY', 'REPRODUCTION CYCLES', 'OTHER'])
+
+# Function to generate random sensor data for a specific cow breed
+def generate_sensor_data(cow_breed):
+    cow_data = {
+        'Holstein Friesian': {'body_temp': (38.0, 39.0), 'rumination': (600, 800), 'steps': (8000, 12000), 'behavior_score': (6, 8)},
+        'Girolando': {'body_temp': (38.0, 39.0), 'rumination': (650, 750), 'steps': (7000, 10000), 'behavior_score': (7, 9)},
+        'Jersey': {'body_temp': (38.0, 39.0), 'rumination': (650, 750), 'steps': (7500, 10500), 'behavior_score': (6, 8)},
+        'Sahiwal': {'body_temp': (38.0, 39.0), 'rumination': (550, 700), 'steps': (6500, 9000), 'behavior_score': (7, 9)},
+        'Kankrej': {'body_temp': (38.0, 39.0), 'rumination': (600, 750), 'steps': (7000, 10000), 'behavior_score': (7, 9)}
+    }
+
+    data = cow_data.get(cow_breed)
+    if not data:
+        return None
+
+    sensor_data = {
+        'Body Temperature (°C)': round(random.uniform(data['body_temp'][0], data['body_temp'][1]), 2),
+        'Rumination (min/day)': random.randint(data['rumination'][0], data['rumination'][1]),
+        'Step Count (steps/day)': random.randint(data['steps'][0], data['steps'][1]),
+        'Behavior Score (1-10)': random.randint(data['behavior_score'][0], data['behavior_score'][1])
+    }
+    return sensor_data
+
+# Load the data and preprocess
+df = load_data()
+
+# Feature Engineering - Create features like daily trends or differences if needed
+df['milk_yield_change'] = df['AGE'].diff().fillna(0)  # Just an example to create a derived feature
+
+# Merge sensor data with cattle profile
+cow_breeds = ['Holstein Friesian', 'Girolando', 'Jersey', 'Sahiwal', 'Kankrej']
+sensor_data_list = []
+for breed in cow_breeds:
+    sensor_data = generate_sensor_data(breed)
+    if sensor_data:
+        sensor_data['TAG ID'] = random.randint(1, 100)  # Random TAG ID for simulation
+        sensor_data['BREED'] = breed
+        sensor_data_list.append(sensor_data)
+
+sensor_df = pd.DataFrame(sensor_data_list)
+
+# Merge with the original cattle profile dataset
+df = pd.merge(df, sensor_df, on='TAG ID', how='left')
+
+# Model Training (Disease Prediction)
+X = df[['Body Temperature (°C)', 'Rumination (min/day)', 'Step Count (steps/day)', 'Behavior Score (1-10)', 'AGE']]
+y = df['OTHER'].apply(lambda x: 1 if x == 'Healthy' else 0)  # Convert health status to binary (0 = not healthy, 1 = healthy)
+
+# Data Preprocessing
+scaler = StandardScaler()
+X_scaled = scaler.fit_transform(X)
+
+# Train Test Split
+X_train, X_test, y_train, y_test = train_test_split(X_scaled, y, test_size=0.3, random_state=42)
+
+# Train a Random Forest Classifier
+model = RandomForestClassifier(n_estimators=100, random_state=42)
+model.fit(X_train, y_train)
+
+# Save the trained model
+joblib.dump(model, 'disease_prediction_model.pkl')
+joblib.dump(scaler, 'scaler.pkl')
+
+# Evaluate the model
+y_pred = model.predict(X_test)
+st.write("Model Evaluation Results")
+st.text(classification_report(y_test, y_pred))
+
+# Streamlit App
+st.title('Cattle Profile & Disease Prediction Dashboard')
+
+# Register New Cow form
+st.header("➕ Register New Cow")
+tag_id = st.number_input("TAG ID", min_value=1, step=1)
+breed = st.selectbox("Breed", cow_breeds)
+age = st.number_input("Age (in years)", min_value=0, step=1)
+surgery = st.text_input("Surgery")
+reproduction_cycles = st.number_input("Reproduction Cycles", min_value=0, step=1)
+other = st.text_input("Other Health Details")
+
+submit = st.button("Register Cow")
+if submit:
+    new_cow = pd.DataFrame([{
+        'TAG ID': tag_id,
+        'BREED': breed,
+        'AGE': age,
+        'SURGERY': surgery,
+        'REPRODUCTION CYCLES': reproduction_cycles,
+        'OTHER': other
+    }])
+    df = pd.concat([df, new_cow], ignore_index=True)
+    st.write("Cow registered successfully!")
+
+# Select RFID dropdown to view details
+st.header("🔢 Select RFID Tag ID")
+selected_rfid = st.selectbox("Select RFID Tag ID", df['TAG ID'].unique())
+
+if selected_rfid:
+    cow_info = df[df['TAG ID'] == selected_rfid].iloc[0]
+    st.write(f"**TAG ID:** {cow_info['TAG ID']}")
+    st.write(f"**BREED:** {cow_info['BREED']}")
+    st.write(f"**AGE:** {cow_info['AGE']} years")
+    st.write(f"**SURGERY:** {cow_info['SURGERY']}")
+    st.write(f"**REPRODUCTION CYCLES:** {cow_info['REPRODUCTION CYCLES']}")
+    st.write(f"**OTHER HEALTH DETAILS:** {cow_info['OTHER']}")
+
+    # Predict disease status based on the selected cow's data
+    selected_data = scaler.transform([[
+        cow_info['Body Temperature (°C)'],
+        cow_info['Rumination (min/day)'],
+        cow_info['Step Count (steps/day)'],
+        cow_info['Behavior Score (1-10)'],
+        cow_info['AGE']
+    ]])
+    disease_prediction = model.predict(selected_data)
+    disease_status = "Healthy" if disease_prediction[0] == 1 else "Not Healthy"
+    st.write(f"Predicted Health Status: {disease_status}")
+
+# Display cow milk yield trends
+st.header("🐄 Cow Milk Yield Trends")
+milk_data = {
+    'Tag ID': df['TAG ID'],
+    'Date': pd.date_range(start='2025-04-01', periods=len(df), freq='D'),
+    'Milk Yield (L)': np.random.uniform(10, 22, len(df)),
+}
+milk_df = pd.DataFrame(milk_data)
+
+# Plot trends
+st.subheader('Milk Yield Trend for Each Cow')
+plt.figure(figsize=(10, 6))
+sns.lineplot(data=milk_df, x='Date', y='Milk Yield (L)', hue='Tag ID', marker='o')
+plt.title('Milk Yield Trend (Day-wise for Each Cow)')
+plt.xlabel('Date')
+plt.ylabel('Milk Yield (L)')
+plt.xticks(rotation=45)
+st.pyplot()
+
+# Save updated data back to CSV
+df.to_csv(DATA_FILE, index=False)