smart-floorplan-predictor 1.0.0

A package for making predictions using a pre-trained ONNX floorplan model

Smart Floorplan Predictor

A Python package for classifying images as EPCs, floorplans, property photos, or exterior shots using a pre-trained ONNX model.

Prerequisites

• Python 3.7 or higher

Installation

pip install smart-floorplan-predictor

Usage

import os
from smart_floorplan_predictor import FloorplanPredictor, FloorplanPredictorError

# --- Configuration for Private Repositories --- 
# If the model is hosted in a private GitHub repository, 
# you MUST provide a GitHub Personal Access Token (PAT).
# Set it as an environment variable:
# export GITHUB_TOKEN="your_github_pat_here"
# Ensure the token has the 'repo' scope to access private repository content.

github_token = os.environ.get("GITHUB_TOKEN")

try:
    # Initialize the predictor
    # If using a private repo, the token will be read from the environment variable
    # or you can pass it directly: FloorplanPredictor(github_token="your_token")
    predictor = FloorplanPredictor(github_token=github_token)

    # Make a prediction using the path to an image file
    image_path = "path/to/your/image.jpg" 
    predicted_class, confidence = predictor.predict_with_confidence(image_path)

    print(f"Predicted Class: {predicted_class}")
    print(f"Confidence: {confidence:.2%}")

    # Or just get the class name directly
    # predicted_class_only = predictor.predict(image_path)
    # print(f"Predicted Class (direct): {predicted_class_only}")

except FloorplanPredictorError as e:
    print(f"An error occurred: {e}")
    # Handle specific errors, e.g., ModelDownloadError

Model Downloading

• The package automatically attempts to download the model.onnx file using the GitHub API if it's not found locally within the package directory (src/smart_floorplan_predictor/). This works even for files stored using Git LFS.
• Private Repositories: Downloading from private GitHub repositories requires a GITHUB_TOKEN environment variable containing a valid Personal Access Token with the repo scope.

Common Errors & Troubleshooting

• ModelDownloadError: ... 404 Not Found ...:
  • Check if the GITHUB_TOKEN environment variable is set correctly (if accessing a private repo).
  • Verify the token is valid, not expired, and has the repo scope enabled in your GitHub Developer settings.
  • Confirm the repository owner, name, and model file path (REPO_OWNER, REPO_NAME, MODEL_FILE_PATH constants in predictor.py) are correct.
• ModelDownloadError: ... 403 Forbidden ...:
  • Usually indicates the provided GITHUB_TOKEN lacks the necessary permissions (repo scope) for a private repository.
• FloorplanPredictorError: GITHUB_TOKEN is required...:
  • You are trying to download from the default private repository location without providing a token. Set the GITHUB_TOKEN environment variable.
• FileNotFoundError or ImageLoadError:
  • Ensure the image path passed to predict or predict_with_confidence is correct and the file exists.
• ONNX Runtime Issues:
  • Ensure onnxruntime is installed correctly for your OS and architecture.

Running Tests (Development)

1. Clone the repository.
2. Install development dependencies: pip install -r requirements.txt (if available) or pip install pytest requests tqdm Pillow numpy onnxruntime.
3. Set the GITHUB_TOKEN environment variable if testing against the private repo.
4. Place test images (like test_image.png) in the root directory.
5. Run the test script: python test.py

Requirements

• Python >= 3.7
• requests
• numpy
• Pillow
• onnxruntime
• tqdm

License

MIT License

Author

Oliver Brown