justingodden/adopted_pets

Will a pet be adopted

Contents

The repo contains the following files:

• data.py: this contains functionality to retrieve task data, and split it into train/val/test datasets.
• engineering.py: this allows for feature engineering of the categorical variables.
• modelling.py: this provides all ML functionality - training, predicting, evaluating, saving, loading.
• main.py: this is the main script which ties all the parts together to complete the task.
• model_dev.ipynb: this Jupyter notebook contains the development of task 1 -- EDA, and model development -- before
• it was turned into python scripts.
• artifacts: this contains the model output of the XGBoost Classifier.
• test_predict.py: contains a single unit test to test the predict function.

Set up

To retrieve the project code, use the following command:

git clone https://github.com/justingodden/adopted_pets.git
cd adopted_pets

Testing

NOTE: it is recommended to run the tests first before running the project code, as the predict function test is set up to work on the current saved ML model file, and running the project code will overwrite the model.

To run the predict function unit test, use the following command in the project root directory:

pytest

Running

To run the main file, use the following command in the root directory:

python main.py

An 'output' folder will be created with a results.csv file inside containing all the predictions generated with the trained XGBoost model.

The model file in the artifacts folder will also be overwritten with each running of the script.