Consider using type hints in function definitions to indicate types of variables and returned data

[alexpreynolds]

epilogos/src/computeEpilogosExpectedMaster.py

Lines 87 to 113 in 80bf1f5

	def s1Calc(file1Path, file2Path, rowsToCalc, numStates, verbose):
	"""
	Function responsible for expected frequency calculation over a set of rows for a saliency metric of 1
	Input:
	file1Path -- The path of the only (single epilogos) or first (paired epilogos) file to read states from
	file2Path -- The path of the second file to read states from (paired epilogos)
	rowsToCalc -- The rows to count expected frequencies from the files
	numStates -- The number of states in the state model
	verbose -- Boolean which if True, causes much more detailed prints
	Output:
	A numpy array containing the counts of each state within the specified rows of the file
	"""
	dataArr = readStates(file1Path=file1Path, file2Path=file2Path, rowsToCalc=rowsToCalc, verbose=verbose)
	expFreqArr = np.zeros(numStates, dtype=np.int32)
	if verbose and rowsToCalc[0] == 0: print("Calculating expected frequencies...", flush=True); tExp = time()
	# Simply count all states across out our subset of data
	uniqueStates, stateCounts = np.unique(dataArr, return_counts=True)
	for i, state in enumerate(uniqueStates):
	expFreqArr[state] += stateCounts[i]
	if verbose and rowsToCalc[0] == 0: print(" Time:", time() - tExp, flush=True)
	return expFreqArr

Type hints can be passed into documentation and linter scripts to automate writing docs and checking the types of data being passed around, including Python primitives (str, int, etc.) and numpy data:

https://www.python.org/dev/peps/pep-0484/

https://numpy.org/devdocs/reference/typing.html

This can also be useful to read the function at a glance, to see what goes in and out, and help with debugging.

[jacobquon]

Will look into this. Definitely seems useful for the final push to Master. Probably will prioritize other things until then