def do_matlab_svm_bdopt (temporal_filtering=True, global_nuis_correction=True):
wd, xd, dd, labelsf, phenof, dataf, masks, dilmasks, templates, pipe = get_filepaths(temporal_filtering, global_nuis_correction)
au.setup_logger(2, logfname=None)
for fi in range(len(dataf)):
myfile = au.remove_ext(subjsf) + '.mat'
au.log.error(exc.message + '\n' + exc.argument)
parser.error(str(msg))
return 0
outdir = args.outdir.strip()
svmdir = args.svmdir.strip()
expname = args.expname.strip()
featsf = args.feats.strip()
testf = args.testf.strip()
cvalue = args.cvalue
lossf = args.lossf
svmargs = args.svmargs.strip()
redoing = args.redoing
verbose = args.verbosity
au.setup_logger(verbose)
#setting command paths
svm_class = 'svm_perf_classify'
svm_learn = 'svm_perf_learn'
if svmdir:
svm_class = svmdir + os.path.sep + svm_class
svm_learn = svmdir + os.path.sep + svm_learn
train_fname = featsf
test_fname = testf
#creating output filename
if expname:
ofname = os.path.basename(expname)
else:
return -1
labelsf = args.labelsf.strip()
dataf = args.dataf.strip()
outdir = args.outdir.strip()
clfmethod = args.estimator.strip()
fsname = args.fsmethod.strip()
prefsmethod = args.prefsmethod.strip()
prefsthr = args.prefsthr
cvfold = args.cvfold.strip()
n_folds = args.nfolds
n_cpus = args.ncpus
verbose = args.verbosity
au.setup_logger(verbose, logfname=None)
scale = True
#label values
n_class = 2
#labels
y = np.loadtxt(labelsf).astype(int)
n_subjs = len(y)
#data
data = np.load(dataf)
n_feats = data.shape[1]
if n_subjs != data.shape[0]:
except argparse.ArgumentError, exc:
print(exc.message + '\n' + exc.argument)
parser.error(str(msg))
return -1
subjlstf = args.subjlstf.strip()
datadir = args.datadir.strip()
feats = args.feats.strip()
outdir = args.outdir.strip()
clfmethod = args.estimator.strip()
fsname = args.fsmethod.strip()
cvfold = args.cvfold.strip()
n_cpus = args.ncpus
verbose = args.verbosity
au.setup_logger(verbose, logfname=None)
scale = True
#label values
n_class = 2
#labels
y, subjs = parse_subjects_list(subjlstf)
scores = np.array(y)
y = np.array(y)
y[scores > 0] = 1
y = y.astype(int)
n_subjs = len(subjs)
maskf = args.mask.strip()
prefix = args.prefix.strip()
leave = args.leave
scale = args.scale
scale_min = args.scale_min
scale_max = args.scale_max
thrps = args.thresholdP.strip().split()
lthr = args.lthreshold.strip()
uthr = args.uthreshold.strip()
absolute = args.absolute
au.setup_logger(args.verbosity)
#checking number of files processed
if not os.path.exists(maskf):
err = 'Mask file not found: ' + maskf
au.log.error(err)
sys.exit(-1)
#number of subjects
subjsnum = au.file_len(subjsf)
#reading subjects list
subjlabels = np.zeros(subjsnum, dtype=int)
subjslist = {}
subjfile = open(subjsf, 'r')
c = 0
from sklearn.lda import LDA
from sklearn.grid_search import GridSearchCV
from sklearn.cross_validation import LeaveOneOut, StratifiedKFold
from sklearn.feature_selection import RFECV
from sklearn.pipeline import Pipeline, FeatureUnion
from IPython.core.debugger import Tracer
debug_here = Tracer()
sys.path.append("/home/alexandre/Dropbox/Documents/phd/work/aizkolari")
import aizkolari_utils as au
"""
import os
import sys
sys.path.append('/home/alexandre/Dropbox/Documents/phd/work/aizkolari')
import aizkolari_utils as au
sys.path.append('/home/alexandre/Dropbox/Documents/phd/work/oasis_aal')
from do_aal_featsets import *
hn = au.get_hostname()
if hn == 'azteca':
wd = '/media/data/oasis_aal'
roisdir = ''
outd = '/media/data/oasis_aal'
elif hn == 'corsair':
#!/usr/bin/python
from IPython.core.debugger import Tracer
debug_here = Tracer()
import os, subprocess, re, sys
import numpy as np
import pickle
sys.path.append('/home/alexandre/Dropbox/Documents/phd/work/aizkolari')
import aizkolari_utils as au
au.setup_logger(verbosity=2)
measures = ['jacs', 'modulatedgm', 'norms', 'trace', 'geodan']
dists = ['pearson', 'bhattacharyya', 'ttest']
study1 = [
'0001', '0002', '0003', '0004', '0005', '0006', '0007', '0008', '0009',
'0010'
]
#studies = {'cv':study1, 'dd':study2}
thrs = [80, 90, 95, 99, 99.5, 99.9, 100]
perfmeas = [
'Accuracy', 'Precision', 'Recall', 'F1', 'PRBEP', 'ROCArea', 'AvgPrec',
'Specificity', 'Brier-score'
]
def do_darya_localizations ():
import scipy.io as sio
sys.path.append('/home/alexandre/Dropbox/Documents/work/visualize_volume')
import visualize_volume as vis
import aizkolari_utils as au
au.setup_logger()
locd = '/home/alexandre/Dropbox/ELM 2013/ELM-2013-Darya/localization'
wd, xd, dd, labelsf, phenof, dataf, masks, dilmasks, templates, pipe = get_filepaths()
maskf = dilmasks[0]
mask, hdr, aff = au.get_nii_data(maskf)
indices = np.array(np.where(mask > 0))
tmpltf = templates[0]
tmplt, _, _ = au.get_nii_data(tmpltf)
flist = os.listdir(locd)
flist = au.find(flist, '.mat')
flist.sort()
for f in flist:
data = sio.loadmat(os.path.join(locd, f))
name = au.remove_ext(f)
if f.find('cobre') >= 0:
p = data['pearson'].squeeze()
locs = p > 0
lidx = indices[:, locs].squeeze()
if f.find('reho') >= 0:
preho = p.copy()
elif f.find('alff') >= 0:
palff = p.copy()
else:
locs = data[name].squeeze()
locs -= 1
if f.find('pearson') >= 0:
if f.find('reho') >= 0:
lidx = indices[:, preho > 0]
elif f.find('alff') >= 0:
lidx = indices[:, palff > 0]
lidx = lidx[:, locs]
else:
lidx = indices[:, locs].squeeze()
locvol = np.zeros_like(mask, dtype=np.float)
locvol[tuple(lidx)] = 1
#save nifti volume
au.save_nibabel (os.path.join(locd, name + '.nii.gz'), locvol, aff, hdr)
#save image
fig = vis.show_many_slices(tmplt, locvol, volaxis=1, vol2_colormap=plt.cm.autumn, figtitle=name)
aizc.save_fig_to_png(fig, os.path.join(locd, name + '.png'))
if os.access (os.path.join(xd,'innercrop'), os.X_OK):
au.exec_command (os.path.join(xd,'innercrop') + ' -o white ' + fig2name + ' ' + fig2name)
#!/usr/bin/python
from IPython.core.debugger import Tracer
debug_here = Tracer()
import os, subprocess, re, sys
import numpy as np
import pickle
sys.path.append("/home/alexandre/Dropbox/Documents/phd/work/aizkolari")
import aizkolari_utils as au
au.setup_logger(verbosity=2)
measures = ["jacs", "modulatedgm", "norms", "trace", "geodan"]
dists = ["pearson", "bhattacharyya", "ttest"]
study1 = ["0001", "0002", "0003", "0004", "0005", "0006", "0007", "0008", "0009", "0010"]
# studies = {'cv':study1, 'dd':study2}
thrs = [80, 90, 95, 99, 99.5, 99.9, 100]
perfmeas = ["Accuracy", "Precision", "Recall", "F1", "PRBEP", "ROCArea", "AvgPrec", "Specificity", "Brier-score"]
cgrid = [10 ** -3, 10 ** -2, 10 ** -1, 10 ** 0, 10 ** 1, 10 ** 2]
scaled = True
# kernel
kernels = ["linear", "rbf"]
def do_oasis_visualize_pca(args):
subjsf = args.infile.strip()
outfile = args.outfile.strip()
datadir = args.datadir.strip()
maskf = args.mask.strip()
fsmethod = args.fsmethod.strip()
# scale = args.scale
# scale_min = args.scale_min
# scale_max = args.scale_max
verbose = args.verbosity
# logging config
au.setup_logger(verbose)
# loading mask
msk = nib.load(maskf).get_data()
nvox = np.sum(msk > 0)
indices = np.where(msk > 0)
# reading subjects list
[scores, subjs] = parse_subjects_list(subjsf, datadir)
scores = np.array(scores)
imgsiz = nib.load(subjs[0]).shape
nsubjs = len(subjs)
# checking mask and first subject dimensions match
if imgsiz != msk.shape:
au.log.error("Subject image and mask dimensions should coincide.")
exit(1)
# relabeling scores to integers, if needed
if not np.all(scores.astype(np.int) == scores):
unis = np.unique(scores)
scs = np.zeros(scores.shape, dtype=int)
for k in np.arange(len(unis)):
scs[scores == unis[k]] = k
y = scs.copy()
else:
y = scores.copy()
# loading data
au.log.info("Loading data...")
X = np.zeros((nsubjs, nvox), dtype="float32")
for f in np.arange(nsubjs):
imf = subjs[f]
au.log.info("Reading " + imf)
img = nib.load(imf).get_data()
X[f, :] = img[msk > 0]
# demo
"""
from sklearn.datasets import fetch_mldata
mnist = fetch_mldata("MNIST original")
X, y = mnist.data[:60000] / 255., mnist.target[:60000]
X, y = shuffle(X, y)
X, y = X[:5000], y[:5000] # lets subsample a bit for a first impression
"""
# lets start plotting
au.log.info("Preparing plots...")
X, y = shuffle(X, y)
X = X / X.max()
# reducing training and test data
if fsmethod != "none":
au.log.info("Feature selecion : " + fsmethod)
selector = select_features(X, y, fsmethod)
X = selector.transform(X)
# au.log.info ('Randomized PCA')
# pca = RandomizedPCA(n_components=2)
au.log.info("Linear Discriminant analysis")
lda = LDA(n_components=2)
fig, plots = plt.subplots(4, 4)
fig.set_size_inches(50, 50)
plt.prism()
for i, j in product(xrange(4), repeat=2):
if i > j:
continue
if i == j:
continue
X_ = X[(y == i) + (y == j)]
y_ = y[(y == i) + (y == j)]
# marks
# marks = y_.astype(str)
# marks[y_ == 0] = 'x'
# marks[y_ == 1] = 'o'
# marks[y_ == 2] = 'D'
# marks[y_ == 3] = '1'
# colors
colors = y_.copy()
colors[y_ == 0] = 0
colors[y_ == 1] = 1
colors[y_ == 2] = 2
colors[y_ == 3] = 3
# transform
# X_trans = pca.fit_transform(X_)
X_trans = lda.fit(X_, y_).transform(X_)
# plots
plots[i, j].scatter(X_trans[:, 0], X_trans[:, 1], c=colors, marker="o")
plots[i, j].set_xticks(())
plots[i, j].set_yticks(())
plots[j, i].scatter(X_trans[:, 0], X_trans[:, 1], c=colors, marker="o")
plots[j, i].set_xticks(())
plots[j, i].set_yticks(())
if i == 0:
plots[i, j].set_title(j)
plots[j, i].set_ylabel(j)
# plt.scatter(X_trans[:, 0], X_trans[:, 1], c=y_)
plt.tight_layout()
plt.savefig(outfile)
