geneset_SV/data_preprocessing.py at main · chiarabales/geneset_SV · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
import numpy as np
import os
import pandas as pd

def create_columns(gen):
    columns = []
    for j in np.arange(gen.shape[0]):
        if j % 50 == 0:
            print(j, len(columns))
        for i in np.arange(2, gen.shape[1]+1):
            if str(gen.loc[j][i]) != 'nan':
                if (gen.loc[j][i] not in columns):
                    columns.append(gen.loc[j][i])
    return columns


def transform_gen(gen, gen_set):

    _data_cols = os.path.join('..', 'data/__COLUMNS')
    name_columns = 'cols_' + gen_set + '.npy'
    cols_where = os.path.join(_data_cols, name_columns)
    cols = np.load(cols_where)

    if gen_set == 'KEGG':
        cols_where = os.path.join(_data_cols, 'cols_KEGG.npy')
        cols = np.load(cols_where)

    elif gen_set == 'CGN':
        cols_where = os.path.join(_data_cols, 'cols_CGN.npy')
        cols = np.load(cols_where)

    elif gen_set == 'CM':
        cols_where = os.path.join(_data_cols, 'cols_CM.npy')
        cols = np.load(cols_where)

    elif gen_set == 'TFT_LEGACY':
        cols_where = os.path.join(_data_cols, 'cols_TFT_LEGACY.npy')
        cols = np.load(cols_where)

    columns = cols.tolist()
    ar = np.zeros(shape = (gen.shape[0],len(columns)))
    data = pd.DataFrame(ar, columns=columns)

    for j in range(gen.shape[0]):
        s = str(gen[0][j])
        data['set'] = 0
        data['set'] = gen[0]
        for i in range(2, gen.shape[1]):
            s = str(gen[i][j])
            if s != 'nan':
                data[s][j] = 1
    return data