def multi_align_stability(ali_params,
mir_stab_thld=0.0,
grp_err_thld=10000.0,
err_thld=1.732,
print_individual=False,
d=64):
def sqerr(a):
n = len(a)
avg = sum(a)
sq = 0.0
for i in range(n):
sq += a[i]**2
return (sq - avg * avg / n) / n
# args - G, data - [T, d]
def func(args, data, return_avg_pixel_error=True):
pass #IMPORTIMPORTIMPORT from math import pi, sin, cos, radians, degrees
ali_params = data[0]
d = data[1]
L = len(ali_params)
N = len(ali_params[0]) / 4
args_list = [0.0] * (L * 3)
for i in range(L * 3 - 3):
args_list[i] = args[i]
cosa = [0.0] * L
sina = [0.0] * L
for i in range(L):
cosa[i] = numpy.cos(numpy.radians(args_list[i * 3]))
sina[i] = numpy.sin(numpy.radians(args_list[i * 3]))
sqr_pixel_error = [0.0] * N
ave_params = []
for i in range(N):
sum_cosa = 0.0
sum_sina = 0.0
sx = [0.0] * L
sy = [0.0] * L
for j in range(L):
if int(ali_params[j][i * 4 + 3]) == 0:
sum_cosa += numpy.cos(
numpy.radians(args_list[j * 3] + ali_params[j][i * 4]))
sum_sina += numpy.sin(
numpy.radians(args_list[j * 3] + ali_params[j][i * 4]))
sx[j] = args_list[j * 3 + 1] + ali_params[j][
i * 4 + 1] * cosa[j] + ali_params[j][i * 4 +
2] * sina[j]
sy[j] = args_list[j * 3 + 2] - ali_params[j][
i * 4 + 1] * sina[j] + ali_params[j][i * 4 +
2] * cosa[j]
else:
sum_cosa += numpy.cos(
numpy.radians(-args_list[j * 3] +
ali_params[j][i * 4]))
sum_sina += numpy.sin(
numpy.radians(-args_list[j * 3] +
ali_params[j][i * 4]))
sx[j] = -args_list[j * 3 + 1] + ali_params[j][
i * 4 + 1] * cosa[j] - ali_params[j][i * 4 +
2] * sina[j]
sy[j] = args_list[j * 3 + 2] + ali_params[j][
i * 4 + 1] * sina[j] + ali_params[j][i * 4 +
2] * cosa[j]
sqrtP = numpy.sqrt(sum_cosa**2 + sum_sina**2)
sqr_pixel_error[i] = max(
0.0, d * d / 4. * (1 - sqrtP / L) + sqerr(sx) + sqerr(sy))
# Get ave transform params
H = EMAN2_cppwrap.Transform({"type": "2D"})
H.set_matrix([
sum_cosa / sqrtP, sum_sina / sqrtP, 0.0,
sum(sx) / L, -sum_sina / sqrtP, sum_cosa / sqrtP, 0.0,
sum(sy) / L, 0.0, 0.0, 1.0, 0.0
])
dd = H.get_params("2D")
# We are using here mirror of the LAST SET.
H = EMAN2_cppwrap.Transform({
"type":
"2D",
"alpha":
dd["alpha"],
"tx":
dd["tx"],
"ty":
dd["ty"],
"mirror":
int(ali_params[L - 1][i * 4 + 3]),
"scale":
1.0
})
dd = H.get_params("2D")
ave_params.append([dd["alpha"], dd["tx"], dd["ty"], dd["mirror"]])
# Warning: Whatever I return here is squared pixel error, this is for the easy expression of derivative
# Don't forget to square root it after getting the value
if return_avg_pixel_error:
return sum(sqr_pixel_error) / N
else:
return sqr_pixel_error, ave_params
"""Multiline Comment0"""
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
#MULTILINEMULTILINEMULTILINE 0
pass #IMPORTIMPORTIMPORT from sp_statistics import k_means_stab_bbenum
pass #IMPORTIMPORTIMPORT from sp_utilities import combine_params2
pass #IMPORTIMPORTIMPORT from numpy import array
pass #IMPORTIMPORTIMPORT from math import sqrt
# I decided not to use scipy in order to reduce the dependency, I wrote the C++ code instead
# from scipy import array, int32
# from scipy.optimize.lbfgsb import fmin_l_bfgs_b
# Find out the subset which is mirror stable over all runs
all_part = []
num_ali = len(ali_params)
nima = len(ali_params[0]) / 4
for i in range(num_ali):
mirror0 = []
mirror1 = []
for j in range(nima):
if ali_params[i][j * 4 + 3] == 0: mirror0.append(j)
else: mirror1.append(j)
mirror0 = numpy.array(mirror0, 'int32')
mirror1 = numpy.array(mirror1, 'int32')
all_part.append([mirror0, mirror1])
match, stab_part, CT_s, CT_t, ST, st = sp_statistics.k_means_stab_bbenum(
all_part, T=0, nguesses=1)
mir_stab_part = stab_part[0] + stab_part[1]
mir_stab_rate = len(mir_stab_part) / float(nima)
if mir_stab_rate <= mir_stab_thld: return [], mir_stab_rate, -1.0
mir_stab_part.sort()
del all_part, match, stab_part, CT_s, CT_t, ST, st
#for i in xrange(len(mir_stab_part)): print i, mir_stab_part[i]
nima2 = len(mir_stab_part)
#print " mirror stable ",nima2
# Keep the alignment parameters of mirror stable particles
ali_params_mir_stab = [[] for i in range(num_ali)]
for i in range(num_ali):
for j in mir_stab_part:
ali_params_mir_stab[i].extend(ali_params[i][j * 4:j * 4 + 4])
# Find out the alignment parameters for each iteration against the last one
args = []
for i in range(num_ali - 1):
alpha, sx, sy, mirror = align_diff_params(
ali_params_mir_stab[i], ali_params_mir_stab[num_ali - 1])
args.extend([alpha, sx, sy])
# Do an initial analysis, purge all outlier particles, whose pixel error are larger than three times the threshold
data = [ali_params_mir_stab, d]
pixel_error_before, ave_params = func(numpy.array(args),
data,
return_avg_pixel_error=False)
# We have to return mir_stab_rate (see above), even if the group does not survive it and the pixel error before cleaning,
# see below, and in ISAC print the user the overall statistics (histograms?)
# so one can get on overall idea how good/bad data is. PAP 01/25/2015
#print " >>> ",sqrt(sum(pixel_error_before)/nima2)
ali_params_cleaned = [[] for i in range(num_ali)]
cleaned_part = []
for j in range(nima2):
pixel_error_before[j] = max(0.0,
pixel_error_before[j]) # prevent sqrt of 0
if numpy.sqrt(pixel_error_before[j]) > 3 * err_thld:
pass #print " removed ",3*err_thld,j,sqrt(pixel_error_before[j])
else:
cleaned_part.append(mir_stab_part[j])
for i in range(num_ali):
ali_params_cleaned[i].extend(
ali_params_mir_stab[i][j * 4:j * 4 + 4])
nima3 = len(cleaned_part)
prever = numpy.sqrt(sum(pixel_error_before) / nima2)
if nima3 <= 1: return [], mir_stab_rate, prever
#print " cleaned part ",nima3
# Use LBFGSB to minimize the sum of pixel errors
data = [ali_params_cleaned, d]
# Use Python code
#ps_lp, val, d = fmin_l_bfgs_b(func, array(args), args=[data], fprime=dfunc, bounds=None, m=10, factr=1e3, pgtol=1e-4, iprint=-1, maxfun=100)
# Use C++ code
ali_params_cleaned_list = []
for params in ali_params_cleaned:
ali_params_cleaned_list.extend(params)
results = EMAN2_cppwrap.Util.multi_align_error(args,
ali_params_cleaned_list, d)
ps_lp = results[:-1]
# Negative val can happen in some rare cases, it should be due to rounding errors,
# because all results show the val is about 1e-13.
#print "Strange results"
#print "args =", args
#print "ali_params_cleaned_list =", ali_params_cleaned_list
#print "results = ", results
val = max(0.0, results[-1])
del ali_params_cleaned_list
if numpy.sqrt(val) > grp_err_thld:
return [], mir_stab_rate, numpy.sqrt(val)
pixel_error_after, ave_params = func(ps_lp,
data,
return_avg_pixel_error=False)
stable_set = []
val = 0.0
for i in range(nima):
if i in cleaned_part:
j = cleaned_part.index(i)
err = numpy.sqrt(pixel_error_after[j])
if err < err_thld:
stable_set.append([err, i, ave_params[j]])
val += err
if print_individual:
sp_global_def.sxprint(
"Particle %4d : pixel error = %18.4f" % (i, err))
else:
if print_individual:
sp_global_def.sxprint(
"Particle %4d : pixel error = %18.4f unstable" %
(i, err))
else:
if print_individual:
sp_global_def.sxprint("Particle %4d : Mirror unstable" % i)
# return average pixel error before pruning as it is more informative
return stable_set, mir_stab_rate, prever # sqrt(val/len(stable_set))
def main():
progname = os.path.basename(sys.argv[0])
usage = progname + " averages1 averages2 --th_grp"
parser = OptionParser(usage, version=SPARXVERSION)
parser.add_option("--T",
type="int",
default=0,
help=" Threshold for matching")
parser.add_option("--J", type="int", default=50, help=" J")
parser.add_option("--max_branching",
type="int",
default=40,
help=" maximum branching")
parser.add_option("--verbose",
action="store_true",
default=False,
help=" Threshold for matching")
parser.add_option("--timing",
action="store_true",
default=False,
help=" Get the timing")
(options, args) = parser.parse_args()
if sp_global_def.CACHE_DISABLE:
from sp_utilities import disable_bdb_cache
disable_bdb_cache()
sp_global_def.BATCH = True
from numpy import array
from sp_statistics import k_means_stab_bbenum
R = len(args)
Parts = []
mem = [0] * R
avg = [0] * R
for r in range(R):
data = EMData.read_images(args[r])
avg[r] = len(data)
part = []
for k in range(len(data)):
lid = data[k].get_attr('members')
mem[r] += len(lid)
lid = array(lid, 'int32')
lid.sort()
part.append(lid.copy())
Parts.append(part)
if options.timing:
from time import time
time1 = time()
MATCH, STB_PART, CT_s, CT_t, ST, st = k_means_stab_bbenum(
Parts,
T=options.T,
J=options.J,
max_branching=options.max_branching,
stmult=0.1,
branchfunc=2)
if options.verbose:
sxprint(MATCH)
sxprint(STB_PART)
sxprint(CT_s)
sxprint(CT_t)
sxprint(ST)
sxprint(st)
sxprint(" ")
for i in range(len(MATCH)):
u = MATCH[i][0] # u is the group in question in partition 1
assert len(STB_PART[u]) == CT_s[u]
sxprint("Group ", end=' ')
for r in range(R):
sxprint("%3d " % (MATCH[i][r]), end=' ')
sxprint(" matches: group size = ", end=' ')
for r in range(R):
sxprint(" %3d" % len(Parts[r][MATCH[i][r]]), end=' ')
sxprint(" matched size = %3d" % (CT_s[u]), end=' ')
if options.verbose:
sxprint(" matched group = %s" % (STB_PART[u]))
else:
sxprint("")
sxprint("\nNumber of averages = ", end=' ')
for r in range(R):
sxprint("%3d" % (avg[r]), end=' ')
sxprint("\nTotal number of particles = ", end=' ')
for r in range(R):
sxprint("%3d" % (mem[r]), end=' ')
sxprint(" number of matched particles = %5d" % (sum(CT_s)))
if options.timing:
sxprint("Elapsed time = ", time() - time1)
sp_global_def.BATCH = False