Makefile

# Makefile
# Eric Tsai, Dec 7, 2022

# Script that downloads all necessary files, processes them, and creates a
# regression model to predict the average rating of a boardgame

# example usage:
# make all

all : doc/boardgame_rating_predictor_report.html

# download the raw ratings file
data/raw/ratings.csv : src/download_data.py
	python src/download_data.py --url="https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2022/2022-01-25/ratings.csv" --out_file="data/raw/ratings.csv"

# download the raw details file
data/raw/details.csv : src/download_data.py
	python src/download_data.py --url="https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2022/2022-01-25/details.csv" --out_file="data/raw/details.csv"

# preprocess the raw files and combine them into one cleaned file
data/processed/boardgame.csv data/processed/training_split.csv data/processed/testing_split.csv : data/raw/ratings.csv data/raw/details.csv src/preprocess_boardgame_data.py
	python src/preprocess_boardgame_data.py --in_file1="data/raw/ratings.csv" --in_file2="data/raw/details.csv" --out_dir="data/processed"

# create and save the figures from EDA
results/numeric_feature_distribution.png results/rating_distribution.png results/top_10_boardgame_categories.png results/top_10_boardgame_families.png results/top_10_boardgame_mechanics.png : data/processed/training_split.csv src/eda_boardgame.py
	python src/eda_boardgame.py --in_file="data/processed/training_split.csv" --out_dir="results/"

# create the regression model by fitting on different models and selecting the best one
results/model_comparison_table.csv results/model_error_score_table.csv results/prediction_comparison_table.csv results/results_plot.png : data/processed/training_split.csv data/processed/testing_split.csv src/prediction_model.py
	python src/prediction_model.py --training_file="data/processed/training_split.csv" --testing_file="data/processed/testing_split.csv" --results_dir="results/"

# render the final report
doc/boardgame_rating_predictor_report.html : doc/boardgame_refs.bib doc/boardgame_rating_predictor_report.Rmd results/numeric_feature_distribution.png results/rating_distribution.png results/top_10_boardgame_categories.png results/top_10_boardgame_families.png results/top_10_boardgame_mechanics.png results/model_comparison_table.csv results/model_error_score_table.csv results/prediction_comparison_table.csv results/results_plot.png
	Rscript -e "rmarkdown::render('doc/boardgame_rating_predictor_report.Rmd')"

	
# delete all generated files and figures
clean: 
	rm -rf data
	rm -rf results