Mercurial > repos > jay > pdaug_word_vector_descriptor
annotate PDAUG_Word_Vector_Descriptor/PDAUG_Word_Vector_Descriptor.py @ 1:a196bbbc7471 draft
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit 3c91f421d26c8f42cf2671e47db735d2cf69dde8"
author | jay |
---|---|
date | Tue, 29 Dec 2020 04:40:30 +0000 |
parents | a3a1d9bea1ad |
children | bfbcebe3eb90 |
rev | line source |
---|---|
0
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
1 import numpy as np |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
2 import os |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
3 import pandas as pd |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
4 from Bio import SeqIO |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
5 from nltk import bigrams |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
6 from nltk import trigrams |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
7 import gensim |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
8 import argparse |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
9 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
10 parser = argparse.ArgumentParser() |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
11 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
12 parser.add_argument("-M", "--ModelInput", required=True, default=None, help="Path to target tsv file") |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
13 parser.add_argument("-R", "--row", required=True, default=None, help="Path to target tsv file") |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
14 parser.add_argument("-I", "--InputFasta", required=True, default=6, help="Path to target tsv file") |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
15 parser.add_argument("-O", "--OutFile", required=False, default='model.txt', help="Path to target tsv file") |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
16 parser.add_argument("-P", "--positive", required=True, help="Path to target tsv file") |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
17 parser.add_argument("-N", "--negative", required=True, help="Path to target tsv file") |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
18 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
19 args = parser.parse_args() |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
20 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
21 seed = 42 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
22 np.random.seed(seed) |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
23 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
24 new_model = gensim.models.KeyedVectors.load_word2vec_format(args.ModelInput, binary=False) |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
25 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
26 import time |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
27 t0 = time.time() |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
28 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
29 temp_word = np.zeros(shape=(int(args.row), 200)) |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
30 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
31 for index, seqs in enumerate(SeqIO.parse(args.InputFasta, 'fasta')): |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
32 seq_sum = 0 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
33 tri_seq = trigrams(seqs.seq) |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
34 for item in ((tri_seq)): |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
35 tri_str = item[0] + item[1] + item[2] |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
36 if tri_str not in list(new_model.wv.vocab): |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
37 continue |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
38 seq_sum = seq_sum + new_model[tri_str] |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
39 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
40 temp_word[index] = seq_sum |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
41 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
42 t1 = time.time() |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
43 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
44 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
45 temp_word = temp_word |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
46 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
47 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
48 clm = [x for x in range(0,temp_word.shape[1])] |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
49 y_temp_word = np.vstack((np.ones((int(args.positive), 1)), np.zeros((int(args.negative),1)))) |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
50 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
51 c, r = y_temp_word.shape |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
52 y_temp_word = y_temp_word.reshape(c,) |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
53 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
54 class_label = pd.DataFrame(y_temp_word, columns=["Class_label"]) |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
55 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
56 df = pd.DataFrame(temp_word, columns=clm) |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
57 df = pd.concat([df,class_label], axis=1) |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
58 |
a3a1d9bea1ad
"planemo upload for repository https://github.com/jaidevjoshi83/pdaug commit a9bd83f6a1afa6338cb6e4358b63ebff5bed155e"
jay
parents:
diff
changeset
|
59 df.to_csv(args.OutFile, index=None, sep="\t") |