def get_madelon_data():
data_dir = os.sep.join(__file__.split(os.sep)[:-3] + ["data"])
filename_dat = data_dir + os.sep + "madelon_train.data"
filename_lab = data_dir + os.sep + "madelon_train.labels"
data = loadtxt(filename_dat)
lab = loadtxt(filename_lab)
return data, lab
def get_madelon_data():
data_dir = os.sep.join(__file__.split(os.sep)[:-3] + ["data"])
filename_dat = data_dir + os.sep + "madelon_train.data"
filename_lab = data_dir + os.sep + "madelon_train.labels"
data = loadtxt(filename_dat)
lab = loadtxt(filename_lab)
return data, lab
def signal(path, skiprows=1):
"""
Loads .ibw or ASCII files and return it as a numpy.ndarray.
Parameters
----------
path : string
Path to signal file.
Returns
-------
signal_array : (n_points, n_signals) array_like
2D real-valued signal array loaded from given .ibw file.
"""
# Get the path and check what the extension is.
ext = splitext(path)[1]
if ext.lower() == '.ibw':
signal_array = loadibw(path)['wave']['wData'] # Load data.
elif ext.lower() == '.txt':
signal_array = loadtxt(path, skiprows=skiprows)
else:
print "Unrecognized file type!"
sys.exit(0)
signal_array.flags.writeable = True # Make array writable.
return signal_array
def get_points(filename):
try:
points = loadtxt(filename)
except:
print "loadtxt failed--reading points manually"
with open(filename, 'r') as f:
lines = [line.strip(' \n').split(' ') for line in f.readlines()]
points = []
for line in lines:
points.append([float(num) for num in line])
points = np.array(points)
return points
def get_glass_data():
data_dir = os.sep.join(__file__.split(os.sep)[:-3] + ["data"])
filename = data_dir + os.sep + "glass.data"
data = loadtxt(filename, delimiter=",")
# create a binary "window glass" vs "non-window glass" labelling
lab = data[:, -1]
lab = array([1. if x <= 4 else -1.0 for x in lab])
# cut off ids and labeling
data = data[:, 1:-1]
return data, lab
def get_pima_data():
data_dir = os.sep.join(__file__.split(os.sep)[:-3] + ["data"])
filename = data_dir + os.sep + "pima-indians-diabetes.data"
data = loadtxt(filename, delimiter=",")
# create labelling
lab = data[:, -1]
lab = array([1. if x == 1 else -1.0 for x in lab])
# cut off labeling
data = data[:, :-1]
return data, lab
def get_glass_data():
data_dir = os.sep.join(__file__.split(os.sep)[:-3] + ["data"])
filename = data_dir + os.sep + "glass.data"
data = loadtxt(filename, delimiter=",")
# create a binary "window glass" vs "non-window glass" labelling
lab = data[:, -1]
lab = array([1. if x <= 4 else -1.0 for x in lab])
# cut off ids and labeling
data = data[:, 1:-1]
return data, lab
def get_pima_data():
data_dir = os.sep.join(__file__.split(os.sep)[:-3] + ["data"])
filename = data_dir + os.sep + "pima-indians-diabetes.data"
data = loadtxt(filename, delimiter=",")
# create labelling
lab = data[:, -1]
lab = array([1. if x == 1 else -1.0 for x in lab])
# cut off labeling
data = data[:, :-1]
return data, lab
def assert_file_matrix(self, filename, M):
try:
with open(filename):
m = loadtxt(filename)
# python loads vectors as 1d-arrays, but we want 2d-col-vectors
if len(shape(m)) == 1:
m = reshape(m, (len(m), 1))
self.assertEqual(M.shape, m.shape)
self.assertLessEqual(norm(m - M), 1e-5)
return True
except IOError:
return False
def __init__ (self, worldfile):
with open (worldfile) as world:
reader = csv.reader (world)
names = {}
for k, line in enumerate (reader):
if not "#" in line [0]:
names [line [0]] = k
data = loadtxt (worldfile, delimiter = ',', usecols = range (1, 7), dtype = float64)
self.names = names
self.radii = data [:, 0]
self.masses = data [:, 1]
self.positions = data [:, 2:4]
self.velocities = data [:, 4:]
self.accelerations = zeros ( (len (self.names),) + self.positions.shape, dtype = float64)
self.mm = outer (self.masses, self.masses)
self.diagind = tuple (range (0, len (self.accelerations)))
self.count = len (self.names)
self.time = 0
def signal(path, skiprows=0):
"""
Loads .ibw or ASCII files and return it as a numpy.ndarray.
:param path: Path to signal file.
:type path: string
:param skiprows:
:type skiprows: int, optional
:returns: 2D real-valued signal array loaded from given .ibw file.
:rtype: (n_points, n_signals) array_like
"""
# Get the path and check what the extension is.
ext = splitext(path)[1]
if ext.lower() == '.ibw':
signal_array = loadibw(path)['wave']['wData'] # Load data.
elif ext.lower() == '.txt':
signal_array = loadtxt(path, skiprows=skiprows)
else:
print("Unrecognized file type!")
sys.exit(0)
try:
signal_array.flags.writeable = True # Make array writable.
except:
pass
return signal_array
from numpy.oldnumeric.random_array import permutation
from matplotlib.pyplot import title, plot, figure, show, draw, clf, contour,\
xlim, ylim, imshow
from numpy.ma.core import arange, mean, reshape, shape, sqrt, floor, zeros
from kameleon_mcmc.kernel.PolynomialKernel import PolynomialKernel
from numpy.linalg.linalg import norm
from numpy.lib.npyio import loadtxt
#from kameleon_mcmc.mcmc.samplers.KameleonWindowLearnScale import KameleonWindowLearnScale
plotting = False
pkernel = PolynomialKernel(degree=3)
samples_long = loadtxt(
"/nfs/home2/dino/kamh-results/StandardMetropolis_PseudoMarginalHyperparameterDistribution_merged_samples.txt"
)
samples_long = samples_long[:10000]
# f_long=open("/nfs/home2/dino/kamh-results/long_experiment_output.bin")
# experiment_long=load(f_long)
# f_long.close()
# thin_long=100
# mcmc_chain_long=experiment_long.mcmc_chain
# burnin=mcmc_chain_long.mcmc_params.burnin
# indices_long = range(burnin, mcmc_chain_long.iteration,thin_long)
# samples_long=mcmc_chain_long.samples[indices_long]
mu_long = mean(samples_long, 0)
print 'using this many samples for the long chain: ', shape(samples_long)[0]
how_many_chains = 20
