An implementation of TANE algorithm to discover function dependences from data.
data.txt: Source data.
fd_discover_using_tuple.ipynb: Implementation using tuple, which costs 24s on data.txt.
fd_discover.ipynb: Implementation using frozenset, which costs 42s on data.txt.
output.txt: Outpout of function dependences.
test_data.txt, test_output.txt: A mini dataset for debugging.