The flight delay prediction task. For more info, please refer to flight_delay.pdf.

• scripts:

  • feat_engineer.py: functions to do the feat_engineer.
  • model.py: the model class to predict.
  • run.py: script to poc model on 9/1 split validation.
  • utli.py: utility functions.
  • get_model.py: run and save model.
  • predict.py: predict by model generated from get_model.py.

• other files:

  • data/airports.dat: airport data from https://openflights.org/data.html.
  • model_history/: model logs.
  • plots/: some plot when EDA.
  • models/: place the model pickle file.
  • flight_delay.pdf: slides for this project.

• Requirements.
  To run the code, please install packages by:

pip install -r requirements.txt

• to run model on training set and validate it.please run the following command.
  • You can set upsampling rate by -u. 0 means no sampling.
  • set -n True if you want to use numeric features.

python run.py 

• to get the model (training on full dataset). Run following command with same arguments as above.

python get_model.py

• to run model on testing set. Run following command.

python predict.py -test_path 'path_to_your_test_file'