Spaces:
Sleeping
Sleeping
File size: 11,057 Bytes
3efedb0 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 |
import pandas as pd
import numpy as np
import pickle
import os
import sys
import logging
from sklearn.preprocessing import StandardScaler
from sklearn.ensemble import RandomForestClassifier
from sklearn.model_selection import train_test_split
from typing import List, Dict, Any
logger = logging.getLogger(__name__)
class LiverDiseaseModel:
def __init__(self):
self.model = None
self.scaler = None
self.feature_names = None
# Get the project root directory
self.project_root = os.path.dirname(os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
# Set paths for model files
self.model_path = os.path.join(self.project_root, 'models', 'liver_disease_model.pkl')
# Default feature names
self.default_feature_names = [
'Age', 'Gender', 'Total_Bilirubin', 'Direct_Bilirubin',
'Alkaline_Phosphotase', 'Alamine_Aminotransferase',
'Aspartate_Aminotransferase', 'Total_Protiens',
'Albumin', 'Albumin_and_Globulin_Ratio'
]
# Initialize feature names
self.feature_names = self.default_feature_names
# Create models directory if it doesn't exist
os.makedirs(os.path.dirname(self.model_path), exist_ok=True)
# Load the model or create a dummy one if not found
self.load_model()
def load_model(self):
"""Load the trained model from disk."""
try:
if os.path.exists(self.model_path):
with open(self.model_path, 'rb') as f:
try:
model_data = pickle.load(f, encoding='latin1')
if isinstance(model_data, dict):
self.model = model_data.get('model')
self.scaler = model_data.get('scaler')
if self.model is None or self.scaler is None:
raise ValueError("Model or scaler missing from loaded data")
else:
self.model = model_data
# Create a new scaler if not found in model data
self.scaler = StandardScaler()
logger.info("Liver disease model loaded successfully")
except Exception as inner_e:
logger.error(f"Error during pickle load: {str(inner_e)}")
raise ValueError(f"Failed to load liver disease model: {str(inner_e)}")
else:
raise FileNotFoundError(f"Liver disease model file not found at {self.model_path}")
except Exception as e:
logger.error(f"Error loading liver disease model: {str(e)}")
raise ValueError(f"Failed to load liver disease model: {str(e)}")
# Remove the _create_dummy_model method entirely
def _create_dummy_model(self):
"""Create a dummy model for testing purposes."""
try:
logger.warning("Creating dummy liver disease model")
self.model = RandomForestClassifier(n_estimators=100, random_state=42)
self.scaler = StandardScaler()
# Create dummy data to fit the scaler and model
dummy_data = pd.DataFrame(np.random.randn(100, len(self.feature_names)),
columns=self.feature_names)
self.scaler.fit(dummy_data)
# Fit the model with dummy data
dummy_target = np.random.randint(0, 2, 100)
self.model.fit(dummy_data, dummy_target)
# Save the dummy model
self.save_model()
logger.info("Dummy liver disease model created and saved successfully")
except Exception as e:
logger.error(f"Error creating dummy liver disease model: {str(e)}")
raise
def save_model(self):
"""Save the model and scaler together in one file."""
try:
# Create a dictionary containing both model and scaler
model_data = {
'model': self.model,
'scaler': self.scaler
}
# Save to file
with open(self.model_path, 'wb') as f:
pickle.dump(model_data, f)
logger.info("Liver disease model and scaler saved successfully")
except Exception as e:
logger.error(f"Error saving liver disease model: {str(e)}")
raise
def predict(self, features: Dict[str, Any]) -> Dict[str, Any]:
"""Make a prediction using the trained model."""
try:
if self.model is None:
raise ValueError(f"Model not loaded. Please ensure model file exists at {self.model_path} and is valid.")
print(f"Input features for liver disease prediction: {features}")
# Convert string inputs to appropriate numeric types
processed_features = {}
for key, value in features.items():
if key == 'Gender':
# Convert 'Male'/'Female' to 1/0
if isinstance(value, str):
processed_features[key] = 1 if value.lower() in ['male', 'm', '1'] else 0
else:
processed_features[key] = 1 if value else 0
else:
# Convert other values to appropriate numeric types
try:
processed_features[key] = float(value)
except (ValueError, TypeError):
# Handle conversion errors
raise ValueError(f"Invalid value for feature {key}: {value}. Expected numeric value.")
# Create DataFrame with processed values
X = pd.DataFrame([processed_features])
# Ensure all required columns are present
for col in self.feature_names:
if col not in X.columns:
raise ValueError(f"Missing required feature: {col}")
# Ensure columns are in the correct order
X = X[self.feature_names]
# Convert all data to float64 to ensure compatibility
X = X.astype(float)
# Scale features
X_scaled = self.scaler.transform(X)
# Make prediction
prediction = bool(self.model.predict(X_scaled)[0])
# Get probability
if hasattr(self.model, 'predict_proba'):
proba = self.model.predict_proba(X_scaled)[0]
probability = float(proba[1]) if len(proba) > 1 else float(proba[0])
else:
probability = 0.5 + (float(self.model.decision_function(X_scaled)[0]) / 10)
probability = max(0, min(1, probability)) # Clamp between 0 and 1
return {
"prediction": prediction,
"probability": probability
}
except Exception as e:
import traceback
traceback.print_exc()
raise ValueError(f"Error during prediction: {str(e)}")
def train_model(self, X, y):
"""Train the model with the given data."""
try:
logger.info("Starting liver disease model training...")
# Initialize the scaler and scale the features
self.scaler = StandardScaler()
X_scaled = self.scaler.fit_transform(X)
# Initialize and train the model
self.model = RandomForestClassifier(
n_estimators=100,
max_depth=10,
random_state=42
)
self.model.fit(X_scaled, y)
# Save the model and scaler
self.save_model()
logger.info("Liver disease model trained successfully")
return True
except Exception as e:
logger.error(f"Error in train_model: {str(e)}")
raise
def get_feature_importance(self):
"""Return feature importance values from the model."""
try:
if self.model is None:
logger.warning("Model not loaded, cannot get feature importance")
return None
# For RandomForestClassifier, we can get feature importance directly
if hasattr(self.model, 'feature_importances_'):
# Return the feature importances as a list
return self.model.feature_importances_.tolist()
else:
# Create dummy feature importance if not available
logger.warning("Feature importance not available in model, returning dummy values")
return [0.15, 0.05, 0.12, 0.08, 0.18, 0.14, 0.10, 0.08, 0.06, 0.04]
except Exception as e:
logger.error(f"Error getting feature importance: {str(e)}")
# Return dummy values as fallback
return [0.15, 0.05, 0.12, 0.08, 0.18, 0.14, 0.10, 0.08, 0.06, 0.04]
def train_model():
"""Train and save the liver disease prediction model"""
try:
model = LiverDiseaseModel()
# Get absolute paths
current_dir = os.path.dirname(os.path.abspath(__file__))
project_root = os.path.dirname(os.path.dirname(current_dir))
data_file = os.path.join(project_root, "data", "indian_liver_patient.csv")
print(f"Loading data from: {data_file}")
print(f"Model will be saved to: {model.model_path}")
# Ensure data file exists
if not os.path.exists(data_file):
raise FileNotFoundError(f"Data file not found at {data_file}")
# Load data
print("Loading and preparing data...")
data = pd.read_csv(data_file)
# Preprocess data
data['Gender'] = data['Gender'].map({'Male': 1, 'Female': 0})
# Handle missing values
data = data.fillna(data.median())
# Select features and target
X = data[model.feature_names]
y = data['Dataset'] # Assuming 'Dataset' is the target column
# Train the model
print("Training model...")
model.train_model(X, y)
print("Model trained and saved successfully")
except Exception as e:
print(f"Error during model training: {str(e)}")
import traceback
print(traceback.format_exc())
sys.exit(1)
if __name__ == "__main__":
train_model() |