MCPcopy Create free account
hub / github.com/PKU-YuanGroup/TaxDiff / decoder_set

Class decoder_set

data_reader/decoder.py:47–90  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

45 super(Uniprot21, self).__init__(chars, encoding=encoding)
46
47class decoder_set:
48 def __iter__(self):
49 return iter(self.x)
50 def __init__(self, path='generate_data/protein/generate1.txt', alphabet=Uniprot21()):
51 print('# loading array:', path, file=sys.stderr)
52
53 sequences = self.load(path, alphabet)
54 # seq_data = []
55 # for i in range(len(sequences)):
56 # x =alphabet.decode(sequences[i].astype('uint8'))
57 # seq_data.append(x.decode())
58 # self.x = seq_data
59 self.x = [[alphabet.decode(seq.astype('uint8')).decode() for seq in sublist] for sublist in sequences]
60 print('# decode', len(self.x), 'sequences', file=sys.stderr)
61
62 def load(self, path, alphabet):
63 data = np.loadtxt(path)
64 data = data - np.ones(data.shape)
65 # sequence = []
66 # star = 0
67 # for i in range(data.shape[0]):
68 # where_zero = np.where(data[i] == -1)
69 # for j in range(np.size(where_zero)):
70 # end = where_zero[0][j]
71 # # sequence length fillter
72 # if end - star > 1:
73 # sequence[i].append(data[i][star:end])
74 # star = where_zero[0][j] + 1
75
76 sequence = [[] for _ in range(data.shape[0])]
77 for i in range(data.shape[0]):
78 star = 0
79 where_zero = np.where(data[i] == -1)[0]
80 for j in range(len(where_zero)):
81 end = where_zero[j]
82 # sequence length filter
83 if end - star > 1:
84 sequence[i].append(data[i][star:end])
85 star = end + 1
86 # Check for a sequence after the last -1
87 if data.shape[1] - star > 1:
88 sequence[i].append(data[i][star:])
89
90 return sequence
91
92def len_select(data):
93 min_length = 10

Callers 1

decode_proteinFunction · 0.85

Calls

no outgoing calls

Tested by

no test coverage detected