m5-forts-castors

/data

raw: CSV files from Kaggle.
interim: raw data merged and formatted (long/grid format). No feature enginnering.
refined: interim data enhanced. It could be a lot of different feature engineering for different algorithms.
exernal: it could be dictionaries containing the hyper-parameters of the algorithms obtained with a Bayesian optimization in the cloud.
submission: outputs files.

python lightgbm_pipeline/raw_data_prep.py --horizon="validation"
python lightgbm_pipeline/raw_data_prep.py --horizon="evaluation"

Name		Name	Last commit message	Last commit date
Latest commit History 126 Commits
.idea		.idea
Hierarchical_Learning_Guillaume		Hierarchical_Learning_Guillaume
data		data
lightgbm_pipeline		lightgbm_pipeline
model_checkpoints		model_checkpoints
prophet_store_dpt_pipeline		prophet_store_dpt_pipeline
scripts		scripts
tf_pipeline		tf_pipeline
wavenet_store_dpt_pipeline		wavenet_store_dpt_pipeline
.gitignore		.gitignore
Error_analysis .ipynb		Error_analysis .ipynb
LICENSE		LICENSE
Makefile		Makefile
README.md		README.md
blend_top_down_weights.ipynb		blend_top_down_weights.ipynb
data_prep.ipynb		data_prep.ipynb
generate_item_level_ensemble_evaluation.ipynb		generate_item_level_ensemble_evaluation.ipynb
generate_item_level_ensemble_validation.ipynb		generate_item_level_ensemble_validation.ipynb
generate_top_down_submission.ipynb		generate_top_down_submission.ipynb
requirements.txt		requirements.txt
score_submission_validation.ipynb		score_submission_validation.ipynb
score_submission_validation.py		score_submission_validation.py
tf_weights.ipynb		tf_weights.ipynb