def pcs_cx(pcs, testing_order, winsize, r, slicing, snr, noise_type):
if snr > 100:
f = open(
"result/cx_testorder%s_hos%d_winsize%d_slice%d_snr%d.csv" %
(testing_order, len(pcs), winsize, slicing, snr), 'w')
for i in range(r):
receive = np.load("temp/data_%d.npy" % (i))[:slicing]
temp = cumx(receive, pcs, len(pcs), testing_order, winsize)
f.write('%s\n' % temp)
print "snr=+inf, ", temp
elif noise_type == "white":
f = open(
"result/cx_testorder%s_hos%d_winsize%d_slice%d_white_snr%d.csv" %
(testing_order, len(pcs), winsize, slicing, snr), 'w')
for i in range(r):
receive = np.load("temp/data_white_%d_%d.npy" % (snr, i))[:slicing]
temp = cumx(receive, pcs, len(pcs), testing_order, winsize)
f.write('%s\n' % temp)
print "white noise, snr=%d, " % (snr), temp
elif noise_type == "color":
f = open(
"result/cx_testorder%s_hos%d_winsize%d_slice%d_color_snr%d.csv" %
(testing_order, len(pcs), winsize, slicing, snr), 'w')
for i in range(r):
receive = np.load("temp/data_color_%d_%d.npy" % (snr, i))[:slicing]
temp = cumx(receive, pcs, len(pcs), testing_order, winsize)
f.write('%s\n' % temp)
print "color noise, snr=%d, " % (snr), temp
else:
print "ERROR: the noise type is wrong!!"
f.close()
def task_cx(pcs, testing_order, winsize, r, slicing, snr, noise_type):
if snr > 100:
f = open("result/cx_testorder%s_hos%d_winsize%d_slice%d_snr%d.csv"%(testing_order, len(pcs), winsize, slicing, snr), 'w')
for i in range(r):
receive = np.load("temp/data_%d.npy"%(i))[:slicing]
temp = cumx(receive, pcs, len(pcs), testing_order, winsize)
f.write('%s\n' % temp)
print "snr=+inf, ", temp
elif noise_type=="white":
f = open("result/cx_testorder%s_hos%d_winsize%d_slice%d_white_snr%d.csv"%(testing_order, len(pcs), winsize, slicing, snr), 'w')
for i in range(r):
receive = np.load("temp/data_white_%d_%d.npy"%(snr, i))[:slicing]
temp = cumx(receive, pcs, len(pcs), testing_order, winsize)
f.write('%s\n' % temp)
print "white noise, snr=%d, "%(snr), temp
elif noise_type=="color":
f = open("result/cx_testorder%s_hos%d_winsize%d_slice%d_color_snr%d.csv"%(testing_order, len(pcs), winsize, slicing, snr), 'w')
for i in range(r):
receive = np.load("temp/data_color_%d_%d.npy"%(snr, i))[:slicing]
temp = cumx(receive, pcs, len(pcs), testing_order, winsize)
f.write('%s\n' % temp)
print "color noise, snr=%d, "%(snr), temp
else:
print "ERROR: the noise type is wrong!!"
f.close()
def maestx(y, pcs, q, norder=3, samp_seg=1, overlap=0):
"""
MAEST MA parameter estimation via the GM-RCLS algorithm, with Tugnait's fix
y - time-series (vector or matrix)
q - MA order
norder - cumulant-order to use [default = 3]
samp_seg - samples per segment for cumulant estimation
[default: length of y]
overlap - percentage overlap of segments [default = 0]
flag - 'biased' or 'unbiased' [default = 'biased']
Return: estimated MA parameter vector
"""
assert norder >= 2 and norder <= 4, "Cumulant order must be 2, 3, or 4!"
nsamp = len(y)
overlap = max(0, min(overlap, 99))
c2 = cumx(y, pcs, 2, q, samp_seg, overlap)
c2 = np.hstack((c2, np.zeros(q)))
cumd = cumx(y, pcs, norder, q, samp_seg, overlap, 0, 0)[::-1]
cumq = cumx(y, pcs, norder, q, samp_seg, overlap, q, q)
cumd = np.hstack((cumd, np.zeros(q)))
cumq[:q] = np.zeros(q)
cmat = toeplitz(cumd, np.hstack((cumd[0], np.zeros(q))))
rmat = toeplitz(c2, np.hstack((c2[0], np.zeros(q))))
amat0 = np.hstack((cmat, -rmat[:, 1:q + 1]))
rvec0 = c2
cumq = np.hstack((cumq[2 * q:q - 1:-1], np.zeros(q)))
cmat4 = toeplitz(cumq, np.hstack((cumq[0], np.zeros(q))))
c3 = cumd[:2 * q + 1]
amat0 = np.vstack((np.hstack((amat0, np.zeros((3*q+1,1)))), \
np.hstack((np.hstack((np.zeros((2*q+1,q+1)), cmat4[:,1:q+1])), \
np.reshape(-c3,(len(c3),1))))))
rvec0 = np.hstack((rvec0, -cmat4[:, 0]))
row_sel = range(q) + range(2 * q + 1, 3 * q + 1) + range(
3 * q + 1, 4 * q + 1) + range(4 * q + 2, 5 * q + 2)
amat0 = amat0[row_sel, :]
rvec0 = rvec0[row_sel]
bvec = lstsq(amat0, rvec0)[0]
b1 = bvec[1:q + 1] / bvec[0]
b2 = bvec[q + 1:2 * q + 1]
if norder == 3:
if all(b2 > 0):
b1 = np.sign(b1) * np.sqrt(0.5 * (b1**2 + b2))
else:
print 'MAEST: alternative solution b1 used'
else:
b1 = np.sign(b2) * (abs(b1) + abs(b2)**(1. / 3)) / 2
# if all(np.sign(b2) == np.sign(b1)):
# b1 = np.sign(b1)* (abs(b1) + abs(b2)**(1./3))/2
# else:
# print 'MAEST: alternative solution b1 used'
return np.hstack(([1], b1))
def arma(pcs, ar, ma, winsize, mc_round, slicing, snr, noise_type):
"""
Input variables:
pcs: array-like pcs choices
ar, ma: order for the ar and ma models
winsize: length of processing segmentation
mc_round: number of monte carlo simulations
slicing: length of slicing for the generated data
snr: snr level for specific data file (larger than 100 for noise free
noice_type: white or color.
"""
if snr > 100:
f1 = open("result/ar%s_hos%d_winsize%d_slice%d_snr%d_pcs%s.csv"%(ar, len(pcs), winsize, slicing, snr, ''.join([str(k) for k in pcs]) ), 'w')
f2 = open("result/ma%s_hos%d_winsize%d_slice%d_snr%d_pcs%s.csv"%(ma, len(pcs), winsize, slicing, snr, ''.join([str(k) for k in pcs]) ), 'w')
for i in range(mc_round):
receive = np.load("temp/arma_data_%d.npy"%(i))[:slicing]
temp = ar_estimate(receive, pcs, ar, ma, winsize)
f1.write('%s\n' % temp)
print "AR - snr=+inf, ", temp
temp = cumx(receive, pcs, len(pcs), ma, winsize)
f2.write('%s\n' % temp)
print "MA - snr=+inf, ", temp
elif noise_type=="white":
f1 = open("result/ar%s_hos%d_winsize%d_slice%d_white_snr%d_pcs%s.csv"%(ar, len(pcs), winsize, slicing, snr, ''.join([str(k) for k in pcs]) ), 'w')
f2 = open("result/ma%s_hos%d_winsize%d_slice%d_white_snr%d_pcs%s.csv"%(ma, len(pcs), winsize, slicing, snr, ''.join([str(k) for k in pcs]) ), 'w')
for i in range(mc_round):
receive = np.load("temp/arma_data_white_%d_%d.npy"%(snr, i))[:slicing]
temp = ar_estimate(receive, pcs, ar, ma, winsize)
f1.write('%s\n' % temp)
print "AR - white noise, snr=%d, "%(snr), temp
temp = cumx(receive, pcs, len(pcs), ma, winsize)
f2.write('%s\n' % temp)
print "MA - white noise, snr=%d, "%(snr), temp
elif noise_type=="color":
f1 = open("result/ar%s_hos%d_winsize%d_slice%d_color_snr%d_pcs%s.csv"%(ar, len(pcs), winsize, slicing, snr, ''.join([str(k) for k in pcs]) ), 'w')
f2 = open("result/ma%s_hos%d_winsize%d_slice%d_color_snr%d_pcs%s.csv"%(ma, len(pcs), winsize, slicing, snr, ''.join([str(k) for k in pcs]) ), 'w')
for i in range(mc_round):
receive = np.load("temp/arma_data_color_%d_%d.npy"%(snr, i))[:slicing]
temp = ar_estimate(receive, pcs, ar, ma, winsize)
f1.write('%s\n' % temp)
print "AR - color noise, snr=%d, "%(snr), temp
temp = cumx(receive, pcs, len(pcs), ma, winsize)
f2.write('%s\n' % temp)
print "MA - color noise, snr=%d, "%(snr), temp
else:
print "ERROR: the noise type is wrong!!"
f1.close()
f2.close()
def task_cx(pcs, taps, winsize, r, slicing):
if len(taps) <= 3:
file_tag = "short"
else:
file_tag = "long"
f = open("cumulant_test_%s_cx%d_%d_%d_slice%d.csv"%(file_tag, len(pcs), winsize, int(''.join(map(str,pcs))), slicing), 'w')
for i in range(r):
signal = np.load("/home/work/rsls/data/exp_deviate_one_%d.npy"%(i))[:slicing]
receive = ir.moving_average(taps, signal)
temp = cx.cumx(receive, pcs, len(pcs), len(taps)-1, winsize)
f.write('%s\n' % temp)
print temp
f.close()
def ar_estimate(sig, pcs, ar, ma, winsize):
if len(pcs) != 3:
raise ValueError("The ar estimate could only handle 3rd-order cumulant")
rb = ma+ar+1
m = np.zeros((rb, 2*ar+ma))
for p in range(rb):
temp = cumx(sig, pcs, 3, 2*ar+ma-1, winsize, 0, p-ar)
m[p,0] = temp[len(temp)/2]
m[p,1:] = (temp[:len(temp)/2][::-1]+temp[len(temp)/2+1:])/2
m = m.T
# put the cumulants into algo. matrix
result = np.zeros((ar*rb, ar))
for i in range(ar*rb):
for j in range(ar):
result[i,j] = m[ma+j+1+i/rb, ar-i%rb]
_, s, _ = np.linalg.svd(result)
return s/s[0]
def ar_estimate(sig, pcs, ar, ma, winsize):
if len(pcs) != 3:
raise ValueError(
"The ar estimate could only handle 3rd-order cumulant")
rb = ma + ar + 1
m = np.zeros((rb, 2 * ar + ma))
for p in range(rb):
temp = cumx(sig, pcs, 3, 2 * ar + ma - 1, winsize, 0, p - ar)
m[p, 0] = temp[len(temp) / 2]
m[p, 1:] = (temp[:len(temp) / 2][::-1] + temp[len(temp) / 2 + 1:]) / 2
m = m.T
# put the cumulants into algo. matrix
result = np.zeros((ar * rb, ar))
for i in range(ar * rb):
for j in range(ar):
result[i, j] = m[ma + j + 1 + i / rb, ar - i % rb]
_, s, _ = np.linalg.svd(result)
return s / s[0]
import maest as ma
import impulse_response as ir
winsize = 512
taps = [1, -2.333, 0.667]
signal = np.load("../data/exp_deviate_one_0.npy")[:10000]
receive = ir.moving_average(taps, signal)
nl = [4, 3, 4]
print "With nested sampling:", ncx.cumx(receive, nl, len(nl),
len(taps) - 1, winsize)
signal = np.load("../data/exp_deviate_one_0.npy")[:10000]
receive = ir.moving_average(taps, signal)
pcs = [2, 3, 5]
print "Without downsampling:", cx.cumx(receive, pcs, len(pcs),
len(taps) - 1, winsize)
signal = np.load("../data/exp_deviate_one_0.npy")[:10000]
receive = ir.moving_average(taps, signal)
pcs = [1, 1, 1]
print "With PCS downsampling:", cx.cumx(receive, pcs, len(pcs),
len(taps) - 1, winsize)
#######################
signal = np.load("../data/exp_deviate_one_0.npy")[:10000]
receive = ir.moving_average(taps, signal)
pcs = [4, 3, 4]
print "With nested sampling:", nma.maestx(receive, pcs,
len(taps) - 1, len(pcs), winsize)
def arma(pcs, ar, ma, winsize, mc_round, slicing, snr, noise_type):
"""
Input variables:
pcs: array-like pcs choices
ar, ma: order for the ar and ma models
winsize: length of processing segmentation
mc_round: number of monte carlo simulations
slicing: length of slicing for the generated data
snr: snr level for specific data file (larger than 100 for noise free
noice_type: white or color.
"""
if snr > 100:
f1 = open(
"result/ar%s_hos%d_winsize%d_slice%d_snr%d_pcs%s.csv" %
(ar, len(pcs), winsize, slicing, snr, ''.join(
[str(k) for k in pcs])), 'w')
f2 = open(
"result/ma%s_hos%d_winsize%d_slice%d_snr%d_pcs%s.csv" %
(ma, len(pcs), winsize, slicing, snr, ''.join(
[str(k) for k in pcs])), 'w')
for i in range(mc_round):
receive = np.load("temp/arma_data_%d.npy" % (i))[:slicing]
temp = ar_estimate(receive, pcs, ar, ma, winsize)
f1.write('%s\n' % temp)
print "AR - snr=+inf, ", temp
temp = cumx(receive, pcs, len(pcs), ma, winsize)
f2.write('%s\n' % temp)
print "MA - snr=+inf, ", temp
elif noise_type == "white":
f1 = open(
"result/ar%s_hos%d_winsize%d_slice%d_white_snr%d_pcs%s.csv" %
(ar, len(pcs), winsize, slicing, snr, ''.join(
[str(k) for k in pcs])), 'w')
f2 = open(
"result/ma%s_hos%d_winsize%d_slice%d_white_snr%d_pcs%s.csv" %
(ma, len(pcs), winsize, slicing, snr, ''.join(
[str(k) for k in pcs])), 'w')
for i in range(mc_round):
receive = np.load("temp/arma_data_white_%d_%d.npy" %
(snr, i))[:slicing]
temp = ar_estimate(receive, pcs, ar, ma, winsize)
f1.write('%s\n' % temp)
print "AR - white noise, snr=%d, " % (snr), temp
temp = cumx(receive, pcs, len(pcs), ma, winsize)
f2.write('%s\n' % temp)
print "MA - white noise, snr=%d, " % (snr), temp
elif noise_type == "color":
f1 = open(
"result/ar%s_hos%d_winsize%d_slice%d_color_snr%d_pcs%s.csv" %
(ar, len(pcs), winsize, slicing, snr, ''.join(
[str(k) for k in pcs])), 'w')
f2 = open(
"result/ma%s_hos%d_winsize%d_slice%d_color_snr%d_pcs%s.csv" %
(ma, len(pcs), winsize, slicing, snr, ''.join(
[str(k) for k in pcs])), 'w')
for i in range(mc_round):
receive = np.load("temp/arma_data_color_%d_%d.npy" %
(snr, i))[:slicing]
temp = ar_estimate(receive, pcs, ar, ma, winsize)
f1.write('%s\n' % temp)
print "AR - color noise, snr=%d, " % (snr), temp
temp = cumx(receive, pcs, len(pcs), ma, winsize)
f2.write('%s\n' % temp)
print "MA - color noise, snr=%d, " % (snr), temp
else:
print "ERROR: the noise type is wrong!!"
f1.close()
f2.close()
import nested_maest as nma
import maest as ma
import impulse_response as ir
winsize = 512
taps = [1, -2.333, 0.667]
signal = np.load("../data/exp_deviate_one_0.npy")[:10000]
receive = ir.moving_average(taps, signal)
nl = [4,3,4]
print "With nested sampling:", ncx.cumx(receive, nl, len(nl), len(taps)-1, winsize)
signal = np.load("../data/exp_deviate_one_0.npy")[:10000]
receive = ir.moving_average(taps, signal)
pcs = [2,3,5]
print "Without downsampling:", cx.cumx(receive, pcs, len(pcs), len(taps)-1, winsize)
signal = np.load("../data/exp_deviate_one_0.npy")[:10000]
receive = ir.moving_average(taps, signal)
pcs = [1,1,1]
print "With PCS downsampling:", cx.cumx(receive, pcs, len(pcs), len(taps)-1, winsize)
#######################
signal = np.load("../data/exp_deviate_one_0.npy")[:10000]
receive = ir.moving_average(taps, signal)
pcs = [4,3,4]
print "With nested sampling:", nma.maestx (receive, pcs, len(taps)-1, len(pcs), winsize)
signal = np.load("../data/exp_deviate_one_0.npy")[:10000]
receive = ir.moving_average(taps, signal)