A critical bug was discovered in the code.

[LitterMa-820]

def transform_data(data,l):
res = []
for i,file in enumerate(data):
tmp = []
res_tmp = []
length = len(file)
for key in file:
bit = bin(key)
bit = bit[2:]
tmp.extend(bit)
for i,_ in enumerate(tmp):
bit = tmp[il:il+l]
bit = ''.join(bit)
bit = bit.zfill(8)
res_tmp.append(bit)
tmp = [int(hex, 2) for hex in tmp]
tmp = tmp[:length]
res.append(tmp)
tmp = []
return res

Why is res_tmp not utilized in the subsequent operations, resulting in the entire code's functionality being reduced to expanding the input binary and then truncating it?

[YHD0225]

During the ablation experiments, the two preprocessing functions were kept separate, and we conducted extensive testing by modifying the variables. However, when merging the two files before uploading them to GitHub, some parameters and variables were incorrectly set. Apologies for the oversight; I have already made the necessary corrections.