def read(self, datadir='./data', file_name='fixed_points.hdf5'):
"""
Read the fixed points from a file.
call signature::
read(self, datadir='./data', file_name='fixed_points.hdf5')
Keyword arguments:
*datadir*:
Directory where the data is stored.
*file_name*:
File with the tracer data.
"""
import numpy as np
# Open the file.
f = h5py.File(os.path.join(datadir, file_name), 'r')
# Extract arrays.
self.t = f['t'].value
self.fidx = f['fidx'].value
self.poincare = f['poincare'].value
# Extract parameters.
params = f['params']
self.params = TracersParameterClass()
for param in params.attrs.keys():
setattr(self.params, param, params.attrs[param])
# Read the time series.
self.fixed_points = []
self.fixed_sign = []
if any(np.array(self.params.int_q) == 'curly_A'):
self.curly_A = []
if any(np.array(self.params.int_q) == 'ee'):
self.ee = []
for t_idx in range(len(self.t)):
group = f['{0}'.format(t_idx)]
self.fixed_points.append(group['fixed_points'].value)
self.fixed_sign.append(group['fixed_sign'].value)
if any(np.array(self.params.int_q) == 'curly_A'):
self.curly_A.append(group['curly_A'].value)
if any(np.array(self.params.int_q) == 'ee'):
self.ee.append(group['ee'].value)
f.close()
# Read the tracers from their own file.
tracer_file_name = self.params.destination[:-5] + '_tracers' + \
self.params.destination[-5:]
self.tracers = Tracers()
self.tracers.read(datadir=datadir, file_name=tracer_file_name)
def read(self, datadir='./data', file_name='fixed_points.hdf5'):
"""
Read the fixed points from a file.
call signature::
read(self, datadir='./data', file_name='fixed_points.hdf5')
Keyword arguments:
*datadir*:
Directory where the data is stored.
*file_name*:
File with the tracer data.
"""
import numpy as np
# Open the file.
f = h5py.File(os.path.join(datadir, file_name), 'r')
# Extract arrays.
self.t = f['t'].value
self.fidx = f['fidx'].value
self.poincare = f['poincare'].value
# Extract parameters.
params = f['params']
self.params = TracersParameterClass()
for param in params.attrs.keys():
setattr(self.params, param, params.attrs[param])
# Read the time series.
self.fixed_points = []
self.fixed_sign = []
if any(np.array(self.params.int_q) == 'curly_A'):
self.curly_A = []
if any(np.array(self.params.int_q) == 'ee'):
self.ee = []
for t_idx in range(len(self.t)):
group = f['{0}'.format(t_idx)]
self.fixed_points.append(group['fixed_points'].value)
self.fixed_sign.append(group['fixed_sign'].value)
if any(np.array(self.params.int_q) == 'curly_A'):
self.curly_A.append(group['curly_A'].value)
if any(np.array(self.params.int_q) == 'ee'):
self.ee.append(group['ee'].value)
f.close()
# Read the tracers from their own file.
tracer_file_name = self.params.destination[:-5] + '_tracers' + \
self.params.destination[-5:]
self.tracers = Tracers()
self.tracers.read(datadir=datadir, file_name=tracer_file_name)
def find_fixed(self,
datadir='data/',
destination='fixed_points.hf5',
varfile='VAR0',
ti=-1,
tf=-1,
trace_field='bb',
h_min=2e-3,
h_max=2e4,
len_max=500,
tol=1e-2,
interpolation='trilinear',
trace_sub=1,
integration='simple',
int_q=[''],
n_proc=1,
tracer_file_name=''):
"""
Find the fixed points.
call signature::
find_fixed(datadir='data/', destination='fixed_points.hf5',
varfile='VAR0', ti=-1, tf=-1, trace_field='bb', h_min=2e-3,
h_max=2e4, len_max=500, tol=1e-2, interpolation='trilinear',
trace_sub=1, integration='simple', int_q=[''], n_proc=1):
Finds the fixed points. Returns the fixed points positions.
Keyword arguments:
*datadir*:
Data directory.
*destination*:
Name of the fixed points file.
*varfile*:
Varfile to be read.
*ti*:
Initial VAR file index for tracer time sequences.
*tf*:
Final VAR file index for tracer time sequences.
*trace_field*:
Vector field used for the streamline tracing.
*h_min*:
Minimum step length for and underflow to occur.
*h_max*:
Parameter for the initial step length.
*len_max*:
Maximum length of the streamline. Integration will stop if
l >= len_max.
*tol*:
Tolerance for each integration step.
Reduces the step length if error >= tol.
*interpolation*:
Interpolation of the vector field.
'mean': takes the mean of the adjacent grid point.
'trilinear': weights the adjacent grid points according to
their distance.
*trace_sub*:
Number of sub-grid cells for the seeds for the initial mapping.
*integration*:
Integration method.
'simple': low order method.
'RK6': Runge-Kutta 6th order.
*int_q*:
Quantities to be integrated along the streamlines.
*n_proc*:
Number of cores for multi core computation.
*tracer_file_name*
Name of the tracer file to be read.
If equal to '' it will compute the tracers.
"""
import numpy as np
# Return the fixed points for a subset of the domain.
def __sub_fixed(queue, ix0, iy0, field, tracers, tidx, var, i_proc):
diff = np.zeros((4, 2))
fixed = []
fixed_sign = []
fidx = 0
poincare_array = np.zeros(
(tracers.x0[i_proc::self.params.n_proc].shape[0],
tracers.x0.shape[1]))
for ix in ix0[i_proc::self.params.n_proc]:
for iy in iy0:
# Compute Poincare index around this cell (!= 0 for potential fixed point).
diff[0, :] = np.array([
tracers.x1[ix, iy, tidx] - tracers.x0[ix, iy, tidx],
tracers.y1[ix, iy, tidx] - tracers.y0[ix, iy, tidx]
])
diff[1, :] = np.array([
tracers.x1[ix + 1, iy, tidx] -
tracers.x0[ix + 1, iy, tidx],
tracers.y1[ix + 1, iy, tidx] -
tracers.y0[ix + 1, iy, tidx]
])
diff[2, :] = np.array([
tracers.x1[ix + 1, iy + 1, tidx] -
tracers.x0[ix + 1, iy + 1, tidx],
tracers.y1[ix + 1, iy + 1, tidx] -
tracers.y0[ix + 1, iy + 1, tidx]
])
diff[3, :] = np.array([
tracers.x1[ix, iy + 1, tidx] -
tracers.x0[ix, iy + 1, tidx],
tracers.y1[ix, iy + 1, tidx] -
tracers.y0[ix, iy + 1, tidx]
])
if sum(np.sum(diff**2, axis=1) != 0):
diff = np.swapaxes(
np.swapaxes(diff, 0, 1) /
np.sqrt(np.sum(diff**2, axis=1)), 0, 1)
poincare = __poincare_index(
field, tracers.x0[ix:ix + 2, iy, tidx],
tracers.y0[ix, iy:iy + 2, tidx], diff)
poincare_array[ix / n_proc, iy] = poincare
if abs(
poincare
) > 5: # Use 5 instead of 2*pi to account for rounding errors.
# Subsample to get starting point for iteration.
nt = 4
xmin = tracers.x0[ix, iy, tidx]
ymin = tracers.y0[ix, iy, tidx]
xmax = tracers.x0[ix + 1, iy, tidx]
ymax = tracers.y0[ix, iy + 1, tidx]
xx = np.zeros((nt**2, 3))
tracers_part = np.zeros((nt**2, 5))
i1 = 0
for j1 in range(nt):
for k1 in range(nt):
xx[i1,
0] = xmin + j1 / (nt - 1.) * (xmax - xmin)
xx[i1,
1] = ymin + k1 / (nt - 1.) * (ymax - ymin)
xx[i1, 2] = self.params.Oz
i1 += 1
for it1 in range(nt**2):
stream = Stream(
field,
self.params,
h_min=self.params.h_min,
h_max=self.params.h_max,
len_max=self.params.len_max,
tol=self.params.tol,
interpolation=self.params.interpolation,
integration=self.params.integration,
xx=xx[it1, :])
tracers_part[it1, 0:2] = xx[it1, 0:2]
tracers_part[it1, 2:] = stream.tracers[
stream.stream_len - 1, :]
min2 = 1e6
minx = xmin
miny = ymin
i1 = 0
for j1 in range(nt):
for k1 in range(nt):
diff2 = (tracers_part[i1+k1*nt, 2] - \
tracers_part[i1+k1*nt, 0])**2 + \
(tracers_part[i1+k1*nt, 3] - \
tracers_part[i1+k1*nt, 1])**2
if diff2 < min2:
min2 = diff2
minx = xmin + j1 / (nt - 1.) * (xmax -
xmin)
miny = ymin + k1 / (nt - 1.) * (ymax -
ymin)
it1 += 1
# Get fixed point from this starting position using Newton's method.
point = np.array([minx, miny])
fixed_point = __null_point(point, var)
# Check if fixed point lies inside the cell.
if ((fixed_point[0] < tracers.x0[ix, iy, tidx]) or
(fixed_point[0] > tracers.x0[ix + 1, iy, tidx])
or (fixed_point[1] < tracers.y0[ix, iy, tidx])
or
(fixed_point[1] > tracers.y0[ix, iy + 1, tidx])):
pass
else:
fixed.append(fixed_point)
fixed_sign.append(np.sign(poincare))
fidx += np.sign(poincare)
queue.put((i_proc, fixed, fixed_sign, fidx, poincare_array))
# Find the Poincare index of this grid cell.
def __poincare_index(field, sx, sy, diff):
poincare = 0
poincare += __edge(field, [sx[0], sx[1]], [sy[0], sy[0]],
diff[0, :], diff[1, :], 0)
poincare += __edge(field, [sx[1], sx[1]], [sy[0], sy[1]],
diff[1, :], diff[2, :], 0)
poincare += __edge(field, [sx[1], sx[0]], [sy[1], sy[1]],
diff[2, :], diff[3, :], 0)
poincare += __edge(field, [sx[0], sx[0]], [sy[1], sy[0]],
diff[3, :], diff[0, :], 0)
return poincare
# Compute rotation along one edge.
def __edge(field, sx, sy, diff1, diff2, rec):
phiMin = np.pi / 8.
dtot = m.atan2(diff1[0] * diff2[1] - diff2[0] * diff1[1],
diff1[0] * diff2[0] + diff1[1] * diff2[1])
if (abs(dtot) > phiMin) and (rec < 4):
xm = 0.5 * (sx[0] + sx[1])
ym = 0.5 * (sy[0] + sy[1])
# Trace the intermediate field line.
stream = Stream(field,
self.params,
h_min=self.params.h_min,
h_max=self.params.h_max,
len_max=self.params.len_max,
tol=self.params.tol,
interpolation=self.params.interpolation,
integration=self.params.integration,
xx=np.array([xm, ym, self.params.Oz]))
stream_x0 = stream.tracers[0, 0]
stream_y0 = stream.tracers[0, 1]
stream_x1 = stream.tracers[stream.stream_len - 1, 0]
stream_y1 = stream.tracers[stream.stream_len - 1, 1]
stream_z1 = stream.tracers[stream.stream_len - 1, 2]
# Discard any streamline which does not converge or hits the boundary.
# if ((stream.len >= len_max) or
# (stream_z1 < self.params.Oz+self.params.Lz-10*self.params.dz)):
# dtot = 0.
if False:
pass
else:
diffm = np.array(
[stream_x1 - stream_x0, stream_y1 - stream_y0])
if sum(diffm**2) != 0:
diffm = diffm / np.sqrt(sum(diffm**2))
dtot = __edge(field, [sx[0], xm], [sy[0], ym], diff1, diffm, rec+1) + \
__edge(field, [xm, sx[1]], [ym, sy[1]], diffm, diff2, rec+1)
return dtot
# Finds the null point of the mapping, i.e. fixed point, using Newton's method.
def __null_point(point, var):
dl = np.min(var.dx, var.dy) / 100.
it = 0
# Tracers used to find the fixed point.
tracers_null = np.zeros((5, 4))
while True:
# Trace field lines at original point and for Jacobian.
# (second order seems to be enough)
xx = np.zeros((5, 3))
xx[0, :] = np.array([point[0], point[1], self.params.Oz])
xx[1, :] = np.array([point[0] - dl, point[1], self.params.Oz])
xx[2, :] = np.array([point[0] + dl, point[1], self.params.Oz])
xx[3, :] = np.array([point[0], point[1] - dl, self.params.Oz])
xx[4, :] = np.array([point[0], point[1] + dl, self.params.Oz])
for it1 in range(5):
stream = Stream(field,
self.params,
h_min=self.params.h_min,
h_max=self.params.h_max,
len_max=self.params.len_max,
tol=self.params.tol,
interpolation=self.params.interpolation,
integration=self.params.integration,
xx=xx[it1, :])
tracers_null[it1, :2] = xx[it1, :2]
tracers_null[it1,
2:] = stream.tracers[stream.stream_len - 1,
0:2]
# Check function convergence.
ff = np.zeros(2)
ff[0] = tracers_null[0, 2] - tracers_null[0, 0]
ff[1] = tracers_null[0, 3] - tracers_null[0, 1]
if sum(abs(ff)) <= 1e-3 * np.min(self.params.dx,
self.params.dy):
fixed_point = np.array([point[0], point[1]])
break
# Compute the Jacobian.
fjac = np.zeros((2, 2))
fjac[0,
0] = ((tracers_null[2, 2] - tracers_null[2, 0]) -
(tracers_null[1, 2] - tracers_null[1, 0])) / 2. / dl
fjac[0,
1] = ((tracers_null[4, 2] - tracers_null[4, 0]) -
(tracers_null[3, 2] - tracers_null[3, 0])) / 2. / dl
fjac[1,
0] = ((tracers_null[2, 3] - tracers_null[2, 1]) -
(tracers_null[1, 3] - tracers_null[1, 1])) / 2. / dl
fjac[1,
1] = ((tracers_null[4, 3] - tracers_null[4, 1]) -
(tracers_null[3, 3] - tracers_null[3, 1])) / 2. / dl
# Invert the Jacobian.
fjin = np.zeros((2, 2))
det = fjac[0, 0] * fjac[1, 1] - fjac[0, 1] * fjac[1, 0]
if abs(det) < dl:
fixed_point = point
break
fjin[0, 0] = fjac[1, 1]
fjin[1, 1] = fjac[0, 0]
fjin[0, 1] = -fjac[0, 1]
fjin[1, 0] = -fjac[1, 0]
fjin = fjin / det
dpoint = np.zeros(2)
dpoint[0] = -fjin[0, 0] * ff[0] - fjin[0, 1] * ff[1]
dpoint[1] = -fjin[1, 0] * ff[0] - fjin[1, 1] * ff[1]
point += dpoint
# Check root convergence.
if sum(abs(dpoint)) < 1e-3 * np.min(self.params.dx,
self.params.dy):
fixed_point = point
break
if it > 20:
fixed_point = point
break
it += 1
return fixed_point
# Find the fixed point using Newton's method, starting at previous fixed point.
def __sub_fixed_series(queue, t_idx, field, var, i_proc):
fixed = []
fixed_sign = []
for i, point in enumerate(
self.fixed_points[t_idx - 1][i_proc::self.params.n_proc]):
fixed_tentative = __null_point(point, var)
# Check if the fixed point lies outside the domain.
if fixed_tentative[0] >= self.params.Ox and \
fixed_tentative[1] >= self.params.Oy and \
fixed_tentative[0] <= self.params.Ox+self.params.Lx and \
fixed_tentative[1] <= self.params.Oy+self.params.Ly:
fixed.append(fixed_tentative)
fixed_sign.append(self.fixed_sign[t_idx - 1][i_proc +
i * n_proc])
queue.put((i_proc, fixed, fixed_sign))
# Discard fixed points which are too close to each other.
def __discard_close_fixed_points(fixed, fixed_sign, var):
fixed_new = []
fixed_sign_new = []
if len(fixed) > 0:
fixed_new.append(fixed[0])
fixed_sign_new.append(fixed_sign[0])
dx = fixed[:, 0] - np.reshape(fixed[:, 0], (fixed.shape[0], 1))
dy = fixed[:, 1] - np.reshape(fixed[:, 1], (fixed.shape[0], 1))
mask = (abs(dx) > var.dx / 2) + (abs(dy) > var.dy / 2)
for idx in range(1, fixed.shape[0]):
if all(mask[idx, :idx]):
fixed_new.append(fixed[idx])
fixed_sign_new.append(fixed_sign[idx])
return np.array(fixed_new), np.array(fixed_sign_new)
# Convert int_q string into list.
if not isinstance(int_q, list):
int_q = [int_q]
self.params.int_q = int_q
if any(np.array(self.params.int_q) == 'curly_A'):
self.curly_A = []
if any(np.array(self.params.int_q) == 'ee'):
self.ee = []
# Multi core setup.
if not (np.isscalar(n_proc)) or (n_proc % 1 != 0):
print("error: invalid processor number")
return -1
queue = mp.Queue()
# Write the tracing parameters.
self.params = TracersParameterClass()
self.params.trace_field = trace_field
self.params.h_min = h_min
self.params.h_max = h_max
self.params.len_max = len_max
self.params.tol = tol
self.params.interpolation = interpolation
self.params.trace_sub = trace_sub
self.params.int_q = int_q
self.params.varfile = varfile
self.params.ti = ti
self.params.tf = tf
self.params.integration = integration
self.params.datadir = datadir
self.params.destination = destination
self.params.n_proc = n_proc
# Make sure to read the var files with the correct magic.
magic = []
if trace_field == 'bb':
magic.append('bb')
if trace_field == 'jj':
magic.append('jj')
if trace_field == 'vort':
magic.append('vort')
if any(np.array(int_q) == 'ee'):
magic.append('bb')
magic.append('jj')
dim = pc.read_dim(datadir=datadir)
# Check if user wants a tracer time series.
if (ti % 1 == 0) and (tf % 1 == 0) and (ti >= 0) and (tf >= ti):
series = True
varfile = 'VAR' + str(ti)
n_times = tf - ti + 1
else:
series = False
n_times = 1
self.t = np.zeros(n_times)
# Read the initial field.
var = pc.read_var(varfile=varfile,
datadir=datadir,
magic=magic,
quiet=True,
trimall=True)
self.t[0] = var.t
grid = pc.read_grid(datadir=datadir, quiet=True, trim=True)
field = getattr(var, trace_field)
param2 = pc.read_param(datadir=datadir, param2=True, quiet=True)
if any(np.array(int_q) == 'ee'):
ee = var.jj * param2.eta - pc.cross(var.uu, var.bb)
# Get the simulation parameters.
self.params.dx = var.dx
self.params.dy = var.dy
self.params.dz = var.dz
self.params.Ox = var.x[0]
self.params.Oy = var.y[0]
self.params.Oz = var.z[0]
self.params.Lx = grid.Lx
self.params.Ly = grid.Ly
self.params.Lz = grid.Lz
self.params.nx = dim.nx
self.params.ny = dim.ny
self.params.nz = dim.nz
tracers = Tracers()
# Create the mapping for all times.
if not tracer_file_name:
tracers.find_tracers(trace_field=trace_field,
h_min=h_min,
h_max=h_max,
len_max=len_max,
tol=tol,
interpolation=interpolation,
trace_sub=trace_sub,
varfile=varfile,
ti=ti,
tf=tf,
integration=integration,
datadir=datadir,
int_q=int_q,
n_proc=n_proc)
else:
tracers.read(datadir=datadir, file_name=tracer_file_name)
self.tracers = tracers
# Set some default values.
self.t = np.zeros((tf - ti + 1) * series + (1 - series))
self.fidx = np.zeros((tf - ti + 1) * series + (1 - series))
self.poincare = np.zeros(
[int(trace_sub * dim.nx),
int(trace_sub * dim.ny), n_times])
ix0 = range(0, int(self.params.nx * trace_sub) - 1)
iy0 = range(0, int(self.params.ny * trace_sub) - 1)
# Start the parallelized fixed point finding.
for tidx in range(n_times):
if tidx > 0:
var = pc.read_var(varfile='VAR' + str(tidx + ti),
datadir=datadir,
magic=magic,
quiet=True,
trimall=True)
field = getattr(var, trace_field)
self.t[tidx] = var.t
proc = []
sub_data = []
fixed = []
fixed_sign = []
for i_proc in range(n_proc):
proc.append(
mp.Process(target=__sub_fixed,
args=(queue, ix0, iy0, field, self.tracers,
tidx, var, i_proc)))
for i_proc in range(n_proc):
proc[i_proc].start()
for i_proc in range(n_proc):
sub_data.append(queue.get())
for i_proc in range(n_proc):
proc[i_proc].join()
for i_proc in range(n_proc):
# Extract the data from the single cores. Mind the order.
sub_proc = sub_data[i_proc][0]
fixed.extend(sub_data[i_proc][1])
fixed_sign.extend(sub_data[i_proc][2])
self.fidx[tidx] += sub_data[i_proc][3]
self.poincare[sub_proc::n_proc, :, tidx] = sub_data[i_proc][4]
for i_proc in range(n_proc):
proc[i_proc].terminate()
# Discard fixed points which lie too close to each other.
fixed, fixed_sign = __discard_close_fixed_points(
np.array(fixed), np.array(fixed_sign), var)
self.fixed_points.append(np.array(fixed))
self.fixed_sign.append(np.array(fixed_sign))
# Compute the traced quantities along the fixed point streamlines.
if any(np.array(self.params.int_q) == 'curly_A') or \
any(np.array(self.params.int_q) == 'ee'):
for t_idx in range(0, n_times):
if any(np.array(self.params.int_q) == 'curly_A'):
self.curly_A.append([])
if any(np.array(self.params.int_q) == 'ee'):
self.ee.append([])
for fixed in self.fixed_points[t_idx]:
# Trace the stream line.
xx = np.array([fixed[0], fixed[1], self.params.Oz])
stream = Stream(field,
self.params,
h_min=self.params.h_min,
h_max=self.params.h_max,
len_max=self.params.len_max,
tol=self.params.tol,
interpolation=self.params.interpolation,
integration=self.params.integration,
xx=xx)
# Do the field line integration.
if any(np.array(self.params.int_q) == 'curly_A'):
curly_A = 0
for l in range(stream.stream_len - 1):
aaInt = vec_int(
(stream.tracers[l + 1] + stream.tracers[l]) /
2,
var,
var.aa,
interpolation=self.params.interpolation)
curly_A += np.dot(
aaInt,
(stream.tracers[l + 1] - stream.tracers[l]))
self.curly_A[-1].append(curly_A)
if any(np.array(self.params.int_q) == 'ee'):
ee_p = 0
for l in range(stream.stream_len - 1):
eeInt = vec_int(
(stream.tracers[l + 1] + stream.tracers[l]) /
2,
var,
ee,
interpolation=self.params.interpolation)
ee_p += np.dot(
eeInt,
(stream.tracers[l + 1] - stream.tracers[l]))
self.ee[-1].append(ee_p)
if any(np.array(self.params.int_q) == 'curly_A'):
self.curly_A[-1] = np.array(self.curly_A[-1])
if any(np.array(self.params.int_q) == 'ee'):
self.ee[-1] = np.array(self.ee[-1])
def find_fixed(self, data_dir='data/', destination='fixed_points.hf5',
varfile='VAR0', ti=-1, tf=-1, trace_field='bb', h_min=2e-3,
h_max=2e4, len_max=500, tol=1e-2, interpolation='trilinear',
trace_sub=1, integration='simple', int_q=[''], n_proc=1):
"""
Find the fixed points.
call signature::
find_fixed(data_dir='data/', destination='fixed_points.hf5',
varfile='VAR0', ti=-1, tf=-1, trace_field='bb', h_min=2e-3,
h_max=2e4, len_max=500, tol=1e-2, interpolation='trilinear',
trace_sub=1, integration='simple', int_q=[''], n_proc=1):
Finds the fixed points. Returns the fixed points positions.
Keyword arguments:
*data_dir*:
Data directory.
*destination*:
Name of the fixed points file.
*varfile*:
Varfile to be read.
*ti*:
Initial VAR file index for tracer time sequences.
*tf*:
Final VAR file index for tracer time sequences.
*trace_field*:
Vector field used for the streamline tracing.
*h_min*:
Minimum step length for and underflow to occur.
*h_max*:
Parameter for the initial step length.
*len_max*:
Maximum length of the streamline. Integration will stop if
l >= len_max.
*tol*:
Tolerance for each integration step.
Reduces the step length if error >= tol.
*interpolation*:
Interpolation of the vector field.
'mean': takes the mean of the adjacent grid point.
'trilinear': weights the adjacent grid points according to
their distance.
*trace_sub*:
Number of sub-grid cells for the seeds for the initial mapping.
*integration*:
Integration method.
'simple': low order method.
'RK6': Runge-Kutta 6th order.
*int_q*:
Quantities to be integrated along the streamlines.
*n_proc*:
Number of cores for multi core computation.
"""
# Return the fixed points for a subset of the domain for the initial
# fixed points.
def __sub_fixed_init(queue, ix0, iy0, field, tracers, var, i_proc):
diff = np.zeros((4, 2))
fixed = []
fixed_sign = []
fidx = 0
poincare_array = np.zeros((tracers.x0[i_proc::self.params.n_proc].shape[0],
tracers.x0.shape[1]))
for ix in ix0[i_proc::self.params.n_proc]:
for iy in iy0:
# Compute Poincare index around this cell (!= 0 for potential fixed point).
diff[0, :] = np.array([tracers.x1[ix, iy, 0] - tracers.x0[ix, iy, 0],
tracers.y1[ix, iy, 0] - tracers.y0[ix, iy, 0]])
diff[1, :] = np.array([tracers.x1[ix+1, iy, 0] - tracers.x0[ix+1, iy, 0],
tracers.y1[ix+1, iy, 0] - tracers.y0[ix+1, iy, 0]])
diff[2, :] = np.array([tracers.x1[ix+1, iy+1, 0] - tracers.x0[ix+1, iy+1, 0],
tracers.y1[ix+1, iy+1, 0] - tracers.y0[ix+1, iy+1, 0]])
diff[3, :] = np.array([tracers.x1[ix, iy+1, 0] - tracers.x0[ix, iy+1, 0],
tracers.y1[ix, iy+1, 0] - tracers.y0[ix, iy+1, 0]])
if sum(np.sum(diff**2, axis=1) != 0):
diff = np.swapaxes(np.swapaxes(diff, 0, 1) /
np.sqrt(np.sum(diff**2, axis=1)), 0, 1)
poincare = __poincare_index(field, tracers.x0[ix:ix+2, iy, 0],
tracers.y0[ix, iy:iy+2, 0], diff)
poincare_array[ix/n_proc, iy] = poincare
if abs(poincare) > 5: # Use 5 instead of 2*pi to account for rounding errors.
# Subsample to get starting point for iteration.
nt = 4
xmin = tracers.x0[ix, iy, 0]
ymin = tracers.y0[ix, iy, 0]
xmax = tracers.x0[ix+1, iy, 0]
ymax = tracers.y0[ix, iy+1, 0]
xx = np.zeros((nt**2, 3))
tracers_part = np.zeros((nt**2, 5))
i1 = 0
for j1 in range(nt):
for k1 in range(nt):
xx[i1, 0] = xmin + j1/(nt-1.)*(xmax-xmin)
xx[i1, 1] = ymin + k1/(nt-1.)*(ymax-ymin)
xx[i1, 2] = self.params.Oz
i1 += 1
for it1 in range(nt**2):
stream = Stream(field, self.params,
h_min=self.params.h_min,
h_max=self.params.h_max,
len_max=self.params.len_max,
tol=self.params.tol,
interpolation=self.params.interpolation,
integration=self.params.integration,
xx=xx[it1, :])
tracers_part[it1, 0:2] = xx[it1, 0:2]
tracers_part[it1, 2:] = stream.tracers[stream.stream_len-1, :]
min2 = 1e6
minx = xmin
miny = ymin
i1 = 0
for j1 in range(nt):
for k1 in range(nt):
diff2 = (tracers_part[i1+k1*nt, 2] - \
tracers_part[i1+k1*nt, 0])**2 + \
(tracers_part[i1+k1*nt, 3] - \
tracers_part[i1+k1*nt, 1])**2
if diff2 < min2:
min2 = diff2
minx = xmin + j1/(nt-1.)*(xmax - xmin)
miny = ymin + k1/(nt-1.)*(ymax - ymin)
it1 += 1
# Get fixed point from this starting position using Newton's method.
point = np.array([minx, miny])
fixed_point = __null_point(point, var)
# Check if fixed point lies inside the cell.
if ((fixed_point[0] < tracers.x0[ix, iy, 0]) or
(fixed_point[0] > tracers.x0[ix+1, iy, 0]) or
(fixed_point[1] < tracers.y0[ix, iy, 0]) or
(fixed_point[1] > tracers.y0[ix, iy+1, 0])):
print "warning: fixed point lies outside the cell"
else:
fixed.append(fixed_point)
fixed_sign.append(np.sign(poincare))
fidx += np.sign(poincare)
queue.put((i_proc, fixed, fixed_sign, fidx, poincare_array))
# Finds the Poincare index of this grid cell.
def __poincare_index(field, sx, sy, diff):
poincare = 0
poincare += __edge(field, [sx[0], sx[1]], [sy[0], sy[0]],
diff[0, :], diff[1, :], 0)
poincare += __edge(field, [sx[1], sx[1]], [sy[0], sy[1]],
diff[1, :], diff[2, :], 0)
poincare += __edge(field, [sx[1], sx[0]], [sy[1], sy[1]],
diff[2, :], diff[3, :], 0)
poincare += __edge(field, [sx[0], sx[0]], [sy[1], sy[0]],
diff[3, :], diff[0, :], 0)
return poincare
# Compute rotation along one edge.
def __edge(field, sx, sy, diff1, diff2, rec):
phiMin = np.pi/8.
dtot = m.atan2(diff1[0]*diff2[1] - diff2[0]*diff1[1],
diff1[0]*diff2[0] + diff1[1]*diff2[1])
if (abs(dtot) > phiMin) and (rec < 4):
xm = 0.5*(sx[0]+sx[1])
ym = 0.5*(sy[0]+sy[1])
# Trace the intermediate field line.
stream = Stream(field, self.params, h_min=self.params.h_min,
h_max=self.params.h_max,
len_max=self.params.len_max,
tol=self.params.tol,
interpolation=self.params.interpolation,
integration=self.params.integration,
xx=np.array([xm, ym, self.params.Oz]))
stream_x0 = stream.tracers[0, 0]
stream_y0 = stream.tracers[0, 1]
stream_x1 = stream.tracers[stream.stream_len-1, 0]
stream_y1 = stream.tracers[stream.stream_len-1, 1]
stream_z1 = stream.tracers[stream.stream_len-1, 2]
# Discard any streamline which does not converge or hits the boundary.
if ((stream.len >= len_max) or
(stream_z1 < self.params.Oz+self.params.Lz-self.params.dz)):
dtot = 0.
else:
diffm = np.array([stream_x1 - stream_x0, stream_y1 - stream_y0])
if sum(diffm**2) != 0:
diffm = diffm/np.sqrt(sum(diffm**2))
dtot = __edge(field, [sx[0], xm], [sy[0], ym], diff1, diffm, rec+1) + \
__edge(field, [xm, sx[1]], [ym, sy[1]], diffm, diff2, rec+1)
return dtot
# Finds the null point of the mapping, i.e. fixed point, using Newton's method.
def __null_point(point, var):
dl = np.min(var.dx, var.dy)/100.
it = 0
# Tracers used to find the fixed point.
tracers_null = np.zeros((5, 4))
while True:
# Trace field lines at original point and for Jacobian.
# (second order seems to be enough)
xx = np.zeros((5, 3))
xx[0, :] = np.array([point[0], point[1], self.params.Oz])
xx[1, :] = np.array([point[0]-dl, point[1], self.params.Oz])
xx[2, :] = np.array([point[0]+dl, point[1], self.params.Oz])
xx[3, :] = np.array([point[0], point[1]-dl, self.params.Oz])
xx[4, :] = np.array([point[0], point[1]+dl, self.params.Oz])
for it1 in range(5):
stream = Stream(field, self.params, h_min=self.params.h_min,
h_max=self.params.h_max, len_max=self.params.len_max,
tol=self.params.tol, interpolation=self.params.interpolation,
integration=self.params.integration, xx=xx[it1, :])
tracers_null[it1, :2] = xx[it1, :2]
tracers_null[it1, 2:] = stream.tracers[stream.stream_len-1, 0:2]
# Check function convergence.
ff = np.zeros(2)
ff[0] = tracers_null[0, 2] - tracers_null[0, 0]
ff[1] = tracers_null[0, 3] - tracers_null[0, 1]
if sum(abs(ff)) <= 1e-3*np.min(self.params.dx, self.params.dy):
fixed_point = np.array([point[0], point[1]])
break
# Compute the Jacobian.
fjac = np.zeros((2, 2))
fjac[0, 0] = ((tracers_null[2, 2] - tracers_null[2, 0]) -
(tracers_null[1, 2] - tracers_null[1, 0]))/2./dl
fjac[0, 1] = ((tracers_null[4, 2] - tracers_null[4, 0]) -
(tracers_null[3, 2] - tracers_null[3, 0]))/2./dl
fjac[1, 0] = ((tracers_null[2, 3] - tracers_null[2, 1]) -
(tracers_null[1, 3] - tracers_null[1, 1]))/2./dl
fjac[1, 1] = ((tracers_null[4, 3] - tracers_null[4, 1]) -
(tracers_null[3, 3] - tracers_null[3, 1]))/2./dl
# Invert the Jacobian.
fjin = np.zeros((2, 2))
det = fjac[0, 0]*fjac[1, 1] - fjac[0, 1]*fjac[1, 0]
if abs(det) < dl:
fixed_point = point
break
fjin[0, 0] = fjac[1, 1]
fjin[1, 1] = fjac[0, 0]
fjin[0, 1] = -fjac[0, 1]
fjin[1, 0] = -fjac[1, 0]
fjin = fjin/det
dpoint = np.zeros(2)
dpoint[0] = -fjin[0, 0]*ff[0] - fjin[0, 1]*ff[1]
dpoint[1] = -fjin[1, 0]*ff[0] - fjin[1, 1]*ff[1]
point += dpoint
# Check root convergence.
if sum(abs(dpoint)) < 1e-3*np.min(self.params.dx, self.params.dy):
fixed_point = point
break
if it > 20:
fixed_point = point
print "warning: Newton did not converged"
break
it += 1
return fixed_point
# Find the fixed point using Newton's method, starting at previous fixed point.
def __sub_fixed_series(queue, t_idx, field, var, i_proc):
fixed = []
fixed_sign = []
for i, point in enumerate(self.fixed_points[t_idx-1][i_proc::self.params.n_proc]):
fixed_tentative = __null_point(point, var)
# Check if the fixed point lies outside the domain.
if fixed_tentative[0] >= self.params.Ox and \
fixed_tentative[1] >= self.params.Oy and \
fixed_tentative[0] <= self.params.Ox+self.params.Lx and \
fixed_tentative[1] <= self.params.Oy+self.params.Ly:
fixed.append(fixed_tentative)
fixed_sign.append(self.fixed_sign[t_idx-1][i_proc+i*n_proc])
queue.put((i_proc, fixed, fixed_sign))
# Discard fixed points which are too close to each other.
def __discard_close_fixed_points(fixed, fixed_sign, var):
fixed_new = []
fixed_new.append(fixed[0])
fixed_sign_new = []
fixed_sign_new.append(fixed_sign[0])
dx = fixed[:, 0] - np.reshape(fixed[:, 0], (fixed.shape[0], 1))
dy = fixed[:, 1] - np.reshape(fixed[:, 1], (fixed.shape[0], 1))
mask = (abs(dx) > var.dx/2) + (abs(dy) > var.dy/2)
for idx in range(1, fixed.shape[0]):
if all(mask[idx, :idx]):
fixed_new.append(fixed[idx])
fixed_sign_new.append(fixed_sign[idx])
return np.array(fixed_new), np.array(fixed_sign_new)
# Convert int_q string into list.
if not isinstance(int_q, list):
int_q = [int_q]
self.params.int_q = int_q
if any(np.array(self.params.int_q) == 'curly_A'):
self.curly_A = []
if any(np.array(self.params.int_q) == 'ee'):
self.ee = []
# Multi core setup.
if not(np.isscalar(n_proc)) or (n_proc%1 != 0):
print "error: invalid processor number"
return -1
queue = mp.Queue()
# Write the tracing parameters.
self.params = TracersParameterClass()
self.params.trace_field = trace_field
self.params.h_min = h_min
self.params.h_max = h_max
self.params.len_max = len_max
self.params.tol = tol
self.params.interpolation = interpolation
self.params.trace_sub = trace_sub
self.params.int_q = int_q
self.params.varfile = varfile
self.params.ti = ti
self.params.tf = tf
self.params.integration = integration
self.params.data_dir = data_dir
self.params.destination = destination
self.params.n_proc = n_proc
# Multi core setup.
if not(np.isscalar(n_proc)) or (n_proc%1 != 0):
print "error: invalid processor number"
return -1
# Make sure to read the var files with the correct magic.
magic = []
if trace_field == 'bb':
magic.append('bb')
if trace_field == 'jj':
magic.append('jj')
if trace_field == 'vort':
magic.append('vort')
if any(np.array(int_q) == 'ee'):
magic.append('bb')
magic.append('jj')
dim = pc.read_dim(datadir=data_dir)
# Check if user wants a tracer time series.
if (ti%1 == 0) and (tf%1 == 0) and (ti >= 0) and (tf >= ti):
series = True
varfile = 'VAR' + str(ti)
n_times = tf-ti+1
else:
series = False
n_times = 1
self.t = np.zeros(n_times)
# Read the initial field.
var = pc.read_var(varfile=varfile, datadir=data_dir, magic=magic,
quiet=True, trimall=True)
self.t[0] = var.t
grid = pc.read_grid(datadir=data_dir, quiet=True, trim=True)
field = getattr(var, trace_field)
param2 = pc.read_param(datadir=data_dir, param2=True, quiet=True)
if any(np.array(int_q) == 'ee'):
ee = var.jj*param2.eta - pc.cross(var.uu, var.bb)
# Get the simulation parameters.
self.params.dx = var.dx
self.params.dy = var.dy
self.params.dz = var.dz
self.params.Ox = var.x[0]
self.params.Oy = var.y[0]
self.params.Oz = var.z[0]
self.params.Lx = grid.Lx
self.params.Ly = grid.Ly
self.params.Lz = grid.Lz
self.params.nx = dim.nx
self.params.ny = dim.ny
self.params.nz = dim.nz
# Create the initial mapping.
tracers = Tracers()
tracers.find_tracers(trace_field='bb', h_min=h_min, h_max=h_max,
len_max=len_max, tol=tol,
interpolation=interpolation,
trace_sub=trace_sub, varfile=varfile,
integration=integration, data_dir=data_dir,
int_q=int_q, n_proc=n_proc)
self.tracers = tracers
# Set some default values.
self.t = np.zeros((tf-ti+1)*series + (1-series))
self.fidx = np.zeros((tf-ti+1)*series + (1-series))
self.poincare = np.zeros([int(trace_sub*dim.nx),
int(trace_sub*dim.ny)])
ix0 = range(0, int(self.params.nx*trace_sub)-1)
iy0 = range(0, int(self.params.ny*trace_sub)-1)
# Start the parallelized fixed point finding for the initial time.
proc = []
sub_data = []
fixed = []
fixed_sign = []
for i_proc in range(n_proc):
proc.append(mp.Process(target=__sub_fixed_init,
args=(queue, ix0, iy0, field, tracers, var,
i_proc)))
for i_proc in range(n_proc):
proc[i_proc].start()
for i_proc in range(n_proc):
sub_data.append(queue.get())
for i_proc in range(n_proc):
proc[i_proc].join()
for i_proc in range(n_proc):
# Extract the data from the single cores. Mind the order.
sub_proc = sub_data[i_proc][0]
fixed.extend(sub_data[i_proc][1])
fixed_sign.extend(sub_data[i_proc][2])
self.fidx[0] += sub_data[i_proc][3]
self.poincare[sub_proc::n_proc, :] = sub_data[i_proc][4]
for i_proc in range(n_proc):
proc[i_proc].terminate()
# Discard fixed points which lie too close to each other.
fixed, fixed_sign = __discard_close_fixed_points(np.array(fixed),
np.array(fixed_sign),
var)
self.fixed_points.append(np.array(fixed))
self.fixed_sign.append(np.array(fixed_sign))
# Find the fixed points for the remaining times.
for t_idx in range(1, n_times):
# Read the data.
varfile = 'VAR' + str(t_idx+ti)
var = pc.read_var(varfile=varfile, datadir=data_dir, magic=magic,
quiet=True, trimall=True)
field = getattr(var, trace_field)
self.t[t_idx] = var.t
# Find the new fixed points.
proc = []
sub_data = []
fixed = []
fixed_sign = []
for i_proc in range(n_proc):
proc.append(mp.Process(target=__sub_fixed_series,
args=(queue, t_idx, field, var, i_proc)))
for i_proc in range(n_proc):
proc[i_proc].start()
for i_proc in range(n_proc):
sub_data.append(queue.get())
for i_proc in range(n_proc):
proc[i_proc].join()
for i_proc in range(n_proc):
# Extract the data from the single cores. Mind the order.
sub_proc = sub_data[i_proc][0]
fixed.extend(sub_data[i_proc][1])
fixed_sign.extend(sub_data[i_proc][2])
for i_proc in range(n_proc):
proc[i_proc].terminate()
# Discard fixed points which lie too close to each other.
fixed, fixed_sign = __discard_close_fixed_points(np.array(fixed),
np.array(fixed_sign),
var)
self.fixed_points.append(np.array(fixed))
self.fixed_sign.append(np.array(fixed_sign))
self.fidx[t_idx] = np.sum(fixed_sign)
# Compute the traced quantities.
if any(np.array(self.params.int_q) == 'curly_A') or \
any(np.array(self.params.int_q) == 'ee'):
for t_idx in range(0, n_times):
if any(np.array(self.params.int_q) == 'curly_A'):
self.curly_A.append([])
if any(np.array(self.params.int_q) == 'ee'):
self.ee.append([])
for fixed in self.fixed_points[t_idx]:
# Trace the stream line.
xx = np.array([fixed[0], fixed[1], self.params.Oz])
stream = Stream(field, self.params,
h_min=self.params.h_min,
h_max=self.params.h_max,
len_max=self.params.len_max,
tol=self.params.tol,
interpolation=self.params.interpolation,
integration=self.params.integration,
xx=xx)
# Do the field line integration.
if any(np.array(self.params.int_q) == 'curly_A'):
curly_A = 0
for l in range(stream.stream_len-1):
aaInt = vec_int((stream.tracers[l+1] + stream.tracers[l])/2,
var, var.aa, interpolation=self.params.interpolation)
curly_A += np.dot(aaInt, (stream.tracers[l+1] - stream.tracers[l]))
self.curly_A[-1].append(curly_A)
if any(np.array(self.params.int_q) == 'ee'):
ee_p = 0
for l in range(stream.stream_len-1):
eeInt = vec_int((stream.tracers[l+1] + stream.tracers[l])/2,
var, ee, interpolation=self.params.interpolation)
ee_p += np.dot(eeInt, (stream.tracers[l+1] - stream.tracers[l]))
self.ee[-1].append(ee_p)
if any(np.array(self.params.int_q) == 'curly_A'):
self.curly_A[-1] = np.array(self.curly_A[-1])
if any(np.array(self.params.int_q) == 'ee'):
self.ee[-1] = np.array(self.ee[-1])