Software Defect Prediction via Attention-Based Recurrent Neural Network

Algorithm 2

Encoding ASTs’ string vectors.

	Input: ASTs’ string vectors , the fixed length of each vector m;
	Output: integer vectors ;
(1)	Initialize a list V, a dict and a dict ;
(2)	for do
(3)	for do
(4)	if not in tokFreq.keys then
(5)	;
(6)	end
(7)	;
(8)	end
(9)	end
(10)	Creating a list sorted in descending order of token frequency which contains tuples of each token and its corresponding frequency;
(11)	for do
(12)	;
(13)	; // establishing a dict of the ordered tokens to map them into integers
(14)	end
(15)	for do
(16)	for do
(17)	;
(18)	end
(19)	if then
(20)	Adding 0s into ;
(21)	end
(22)	else if then
(23)	for do
(24)	; // finding the index of the lowest token frequency ; // deleting the token
(25)	end
(26)	end
(27)	Adding into V;
(28)	end
(29)	return V;