-
Notifications
You must be signed in to change notification settings - Fork 0
/
preparedata.py
133 lines (92 loc) · 3.96 KB
/
preparedata.py
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
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Wed Dec 14 19:37:30 2016
@author: ecr05
"""
from collections import namedtuple
import mapping as cm
import numpy as np
import os
import nibabel
import glob
def get_idlist(regularexpression):
"""
get subject ids from files
Parameters
----------
regularexpression : regular expression for data files
Returns
-------
idlist : list of subject ids
"""
filenames=glob.glob(regularexpression)
ids=[idn.replace('/data/PROJECTS/HCP/HCP_PARCELLATION/TRAININGDATA/featuresets/', '') for idn in filenames]
ids=[idn.replace('.L.MultiModal_Features_MSMAll_2_d41_WRN_DeDrift.FULLVISUO.32k_fs_LR.func.gii', '') for idn in ids]
return ids;
def get_datalists(use_labels, normalise,paths):
"""
load all original gifti data
Parameters
----------
paths : dictionary with data paths
Returns
-------
dataset : named tuple containing feature data
labelset: named tuple containing label data
"""
datalist=paths['list']
DataMatrix = namedtuple('DataMatrix','DATA,ids,samples,features') # namedtupele for holding and indexing all the data
trainingfunc = nibabel.load(paths['fname'].replace('%subjid%',datalist[1]))
numfeatures = trainingfunc.numDA
numdatapoints = trainingfunc.darrays[0].data.shape[0];
# create named tuple for data files
dataset = DataMatrix(DATA=np.zeros((numdatapoints,(len(datalist)*numfeatures))),ids=[],samples=len(datalist),features=numfeatures)
start=0
if use_labels:
# create datamatrix for all label files
labelset = np.zeros((numdatapoints,len(datalist)))
for ind,name in enumerate(datalist):
print(name,ind,len(datalist))
func = nibabel.load(paths['fname'].replace('%subjid%',name))
if use_labels:
# fill label array with single subject labels
label = nibabel.load(paths['lname'].replace('%subjid%',name))
labelset[:,ind] = label.darrays[0].data
for d in range(0,func.numDA):
#fill data array
# =============================================================================
# if normalise:
# subjdata_d=cm.normalize(func.darrays[d].data,0)
# else:
# subjdata_d=func.darrays[d].data
# =============================================================================
dataset.DATA[:,start+d] =func.darrays[d].data# subjdata_d
dataset.ids.append(name)
start += numfeatures
ALLDATA={}
ALLDATA['data']=dataset
if use_labels:
ALLDATA['labels']=labelset
return ALLDATA
def project_data(DATA, interpolator,data_paths,resampleH,resampleW,use_labels,normalise):
"""
project data from sphere to plane
Parameters
----------
data_set : struct containing single subject featuresets
interpolator : links spherical mesh grid points to coordinates on 2D project
(currently only nearest neighbour available)
paths : dictionary with data paths
Returns
-------
"""
cm.project_data(DATA, interpolator, data_paths, resampleH, resampleW,normalise)
# =============================================================================
# if paths.usegrouplabels:
# filename = os.path.join(Odir, 'projections'+ aug+n_end)
# else:
# filename = os.path.join(Odir, 'projections'+ aug+'Nature' +n_end)
#
# =============================================================================
cm.write_projection_paths(DATA['data'], os.path.join(data_paths['Odirname'],data_paths['abr']+'.pk1'), data_paths,use_labels,normalise)