def A_from_structure(structure_filepath):
"""User defined function, to build the A matrices.
Parameters
----------
structure : string, filepath to the structure
Return
------
A, structure
Those two objects will be accessible via global variables: A and STRUCTURE
"""
sys.path.append(
os.path.join(os.getenv('HOME'), 'gits', 'scripts',
'2013_brainomics_genomics'))
from bgutils.build_websters import group_pw_snp2
#a terme on passe ici la fichier vers les données de contrainte
group, group_names, snpList = group_pw_snp2(fic='go_synaptic_snps_gene',
cache=True)
tmp = []
for i in group:
tmp.extend(group[i])
p = len(set(tmp))
print "DEBUG: ", p
import parsimony.functions.nesterov.gl as gl
weights = [np.sqrt(len(group[i])) for i in group]
A = gl.A_from_groups(p, groups=group, weights=weights)
structure = None
return A, structure
def load_globals(config):
import mapreduce as GLOBAL # access to global variables
GLOBAL.DATA = GLOBAL.load_data(config["data"])
from bgutils.build_websters import group_pw_snp2
fic = 'go_synaptic_snps_gene' #'go_synaptic_snps_gene10'
groups, group_names, snpList = group_pw_snp2(fic=fic, cache=True)
GLOBAL.groups = groups
def mapreduce_parameterize():
# n_samples, shape = 200, (100, 100, 1)
# base = "classif_%ix%ix%ix%i_" % tuple([n_samples] + list(shape))
# X3d, y, beta3d, proba = datasets.classification.dice5.load(n_samples=n_samples,
# shape=shape, snr=5, random_seed=1)
# X = X3d.reshape((n_samples, np.prod(beta3d.shape)))
base = 'grid'
# 1- read constraints : we do not use Group Constraint here
sys.path.append(
os.path.join(os.getenv('HOME'), 'gits', 'scripts',
'2013_brainomics_genomics'))
from bgutils.build_websters import group_pw_snp2, get_websters_logr, pw_gene_snp2
group, group_names, snpList = group_pw_snp2(fic='go_synaptic_snps_gene10',
cache=True)
pw, _ = pw_gene_snp2(fic='go_synaptic_snps_gene10', cache=True)
# 2- get the snps list to get a data set w/ y continous variable
# convenient snp order
# subject order granted by the method
snp_subset = np.asarray(snpList, dtype=str).tolist()
y, X = get_websters_logr(snp_subset=snp_subset)
#3 ajout regresseur 1
X = np.hstack((np.ones((X.shape[0], 1)), X))
# create set of train test
cv = StratifiedKFold(y.ravel(), n_folds=5)
cv = [[tr.tolist(), te.tolist()] for tr, te in cv]
# parameters grid
alphas = [.01, .05, .1, .5, 1.]
# couplage_kl_g = np.array([[1., 0., 1], [0., 1., 1], [.1, .9, 1], [.9, .1, 1], [.5, .5, 1]])
couplage_kl_g = np.array([[1., 0., 1], [0., 0., 1], [.1, 0., 1],
[.9, 0., 1], [.5, 0., 1]])
# k_range = np.arange(0, 1., .1)
k_range = np.arange(0, 3., .5)
alphas = [.01, .05, .1, .5, 1.]
l2l1k = [
np.array([[float(1 - k), float(1 - k), k]]) * couplage_kl_g
for k in k_range
]
l2l1k.append(np.array([[0., 0., 1.]]))
l2l1k = np.concatenate(l2l1k)
alphal2l1k = np.concatenate([
np.c_[np.array([[alpha]] * l2l1k.shape[0]), l2l1k] for alpha in alphas
])
params = [params.tolist() for params in alphal2l1k]
# Save X, y, mask structure and cv
np.save(base + 'X.npy', X)
np.save(base + 'y.npy', y)
# nibabel.Nifti1Image(np.ones(shape), np.eye(4)).to_filename(base+'mask.nii')
# Save everything in a single file config.json
config = dict(
data="%s?.npy" % base,
structure=base + 'mask.nii',
params=params,
resample=cv,
map_output=base + "map_results",
job_file=base + "jobs.json",
user_func=
"/home/vf140245/gits/scripts/2013_brainomics_genomics/scripts/cv_webs_logr_ridge_gl.py",
cores=8,
reduce_input=base + "map_results/*/*",
reduce_group_by=base + "map_results/.*/(.*)")
o = open("config.json", "w")
json.dump(config, o)
o.close()
#"""
#run /home/hl237680/gits/scripts/2013_imagen_bmi/scripts/15_cv_multivariate_residualized_BMI.py
#"""
if __name__ == "__main__":
## Set pathes
#WD = "/home/fh235918/git/scripts/2013_brainomics_genomics/test"
WD = OUTPUT_DIR
if not os.path.exists(WD): os.makedirs(WD)
print '#############'
print '# Read data #'
print '#############'
fic = 'go_synaptic_snps_gene' #'go_synaptic_snps_gene10'
groups, group_names, snpList = group_pw_snp2(fic=fic, cache=True)
# 2- get the snps list to get a data set w/ y continous variable
# convenient snp order
# subject order granted by the method
y, X_orig = get_websters_logr(snp_subset=snpList.tolist())
# 3- fix X : add a ones constant regressor
n, p = X_orig.shape # keep orig size
X = np.hstack((np.ones((X_orig.shape[0], 1)), X_orig)) # add intercept
print "#####################"
print "# Build config file #"
print "#####################"
## Parameterize the mapreduce
## 1) pathes
@author: vf140245
Copyrignt : CEA NeuroSpin - 2014
"""
import sys, os
import numpy as np
import time
import matplotlib.pyplot as plt
sys.path.append(os.path.join(os.getenv('HOME'),
'gits','scripts','2013_brainomics_genomics'))
if __name__=="__main__":
# read constraints : we do not use Group Constraint here
from bgutils.build_websters import group_pw_snp2,get_websters_linr, pw_gene_snp2
group, group_names, snpList = group_pw_snp2(fic='go_synaptic_snps_gene10')
pw, _ = pw_gene_snp2(fic='go_synaptic_snps_gene10')
# get the snps list to get a data set w/ y continous variable
# convenient snp order
# subject order granted by the method
snp_subset=np.asarray(snpList,dtype=str).tolist()
y, X = get_websters_linr(snp_subset=snp_subset)
# fix X : add a ones constant regressor
p = (X.shape)[1] # keep orig size
X = np.hstack((np.ones((X.shape[0],1)),X)) # add intercept
# build A matrix
import parsimony.functions.nesterov.gl as gl
weights = [np.sqrt(len(group[i])) for i in group]
x = transcode(chr_plot_info[k], x)
y = beta[indices]
# if k == u'chr3':
# print indices
# print x
# print y
plt.vlines(x, [0], y, lw=2)
plt.subplots_adjust(left=0.01, right=0.99, top=0.99, bottom=0.01)
plt.show()
def get_gene_from_pw(pw):
gene_list = []
for i in pw:
for j in pw[i]:
gene_list.append(j)
return np.unique(gene_list).tolist()
if __name__ == "__main__":
# read constraints : we do not use Group Constraint here
from bgutils.build_websters import group_pw_snp2, pw_gene_snp2
cache = True
group, group_names, snpList = group_pw_snp2(nb=10, cache=cache)
pw, _ = pw_gene_snp2(nb=10, cache=cache)
beta = np.random.uniform(-1, 1, 539)
plot_pw(beta, pway=pw, snplist=snpList)
