Spaces:
Runtime error
Runtime error
Upload app.py with huggingface_hub
Browse files
app.py
ADDED
|
@@ -0,0 +1,98 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
from flask import Flask, request, jsonify
|
| 2 |
+
import joblib
|
| 3 |
+
import numpy as np
|
| 4 |
+
import pandas as pd
|
| 5 |
+
import os
|
| 6 |
+
|
| 7 |
+
# Create a Flask application instance
|
| 8 |
+
app = Flask(__name__)
|
| 9 |
+
|
| 10 |
+
# Define the path to the serialized model file
|
| 11 |
+
model_filename = 'best_model.joblib'
|
| 12 |
+
model_path = os.path.join(os.path.dirname(__file__), model_filename)
|
| 13 |
+
|
| 14 |
+
|
| 15 |
+
# Load the serialized model
|
| 16 |
+
try:
|
| 17 |
+
loaded_model = joblib.load(model_path)
|
| 18 |
+
print(f"Model loaded successfully from '{model_path}'")
|
| 19 |
+
except Exception as e:
|
| 20 |
+
print(f"Error loading model: {e}")
|
| 21 |
+
loaded_model = None # Set model to None if loading fails
|
| 22 |
+
|
| 23 |
+
# Define the prediction endpoint
|
| 24 |
+
@app.route('/predict', methods=['POST'])
|
| 25 |
+
def predict():
|
| 26 |
+
if loaded_model is None:
|
| 27 |
+
return jsonify({'error': 'Model not loaded'}), 500
|
| 28 |
+
|
| 29 |
+
try:
|
| 30 |
+
# Get the data from the request
|
| 31 |
+
data = request.get_json()
|
| 32 |
+
|
| 33 |
+
# Convert the input data to a pandas DataFrame
|
| 34 |
+
# Assuming the input data is a list of dictionaries,
|
| 35 |
+
# where each dictionary represents a data point with features.
|
| 36 |
+
# The order of features should match the training data.
|
| 37 |
+
input_df = pd.DataFrame(data)
|
| 38 |
+
|
| 39 |
+
# NOTE: The preprocessor object is not available here.
|
| 40 |
+
# In a real deployment, you would also serialize and load the preprocessor
|
| 41 |
+
# or recreate it with the exact same steps and parameters.
|
| 42 |
+
# For this example, we'll assume the input is already preprocessed or
|
| 43 |
+
# we skip preprocessing for simplicity (not recommended for production).
|
| 44 |
+
# If preprocessing is needed, you would do:
|
| 45 |
+
# input_processed = loaded_preprocessor.transform(input_df)
|
| 46 |
+
# For this example, let's assume the input data is already in the expected
|
| 47 |
+
# format for the loaded model (which expects processed features).
|
| 48 |
+
# In a real scenario, ensure the input data format matches the training data.
|
| 49 |
+
# Assuming the input data is already preprocessed (e.g., one-hot encoded and scaled)
|
| 50 |
+
# and is in the form of a list of lists or a numpy array that can be
|
| 51 |
+
# converted to a format compatible with the loaded model's expected input shape.
|
| 52 |
+
# For simplicity in this example, we will assume the input JSON
|
| 53 |
+
# directly corresponds to the processed features expected by the model.
|
| 54 |
+
# In a real-world scenario, you would need to implement the preprocessing steps here
|
| 55 |
+
# using the serialized preprocessor or by recreating it.
|
| 56 |
+
|
| 57 |
+
# Convert input_df to numpy array or appropriate format if needed by the model
|
| 58 |
+
# Based on the training code, the model was trained on a processed numpy array
|
| 59 |
+
# We will assume the input data JSON is structured to represent this processed array.
|
| 60 |
+
# If the input JSON is raw data, you'll need the preprocessor here.
|
| 61 |
+
# For now, let's assume the input JSON data can be directly passed to predict
|
| 62 |
+
# if it's structured correctly as a list of lists or similar.
|
| 63 |
+
# A safer approach would be to expect raw data and use a loaded preprocessor.
|
| 64 |
+
|
| 65 |
+
# For now, let's assume the input data JSON is a list of dictionaries,
|
| 66 |
+
# and we convert it to a DataFrame and then to a numpy array for prediction
|
| 67 |
+
# if the model expects a numpy array.
|
| 68 |
+
# However, since the model was trained on X_processed which was a sparse matrix initially
|
| 69 |
+
# from the ColumnTransformer, direct conversion to a numpy array might lose sparsity
|
| 70 |
+
# or cause issues if the original preprocessor's output format is critical.
|
| 71 |
+
# A robust solution requires serializing and loading the preprocessor as well.
|
| 72 |
+
|
| 73 |
+
# Given the context of previous cells, X_processed was likely a numpy array after transformation.
|
| 74 |
+
# Let's assume the input JSON data can be directly used to create a numpy array
|
| 75 |
+
# with the correct number of features (34 in this case, based on X_processed shape).
|
| 76 |
+
|
| 77 |
+
# Example assuming input data is a list of lists matching the processed features shape
|
| 78 |
+
input_data_processed = np.array(data['features'])
|
| 79 |
+
|
| 80 |
+
|
| 81 |
+
# Make predictions
|
| 82 |
+
predictions = loaded_model.predict(input_data_processed)
|
| 83 |
+
|
| 84 |
+
# Convert predictions to a list
|
| 85 |
+
predictions_list = predictions.tolist()
|
| 86 |
+
|
| 87 |
+
# Return the predictions as a JSON response
|
| 88 |
+
return jsonify({'predictions': predictions_list})
|
| 89 |
+
|
| 90 |
+
except Exception as e:
|
| 91 |
+
return jsonify({'error': str(e)}), 400
|
| 92 |
+
|
| 93 |
+
# To run the Flask application (for local testing)
|
| 94 |
+
if __name__ == '__main__':
|
| 95 |
+
# For local testing, you can run:
|
| 96 |
+
# Ensure the model file is in the same directory or adjust the model_path
|
| 97 |
+
# app.run(debug=True)
|
| 98 |
+
pass
|