def post_compute(variables = ['b2m'], datadir = 'data'):
"""
Compute diagnostic variables from the VAR files.
call signature::
post_compute(variables = ['urms'], datadir = 'data')
Read the VAR files and compute the diagnostic variables.
Write the result in 'post_evaluation.dat'.
Keyword arguments:
*variables*:
The diagnostic variables to be computed.
*datadir*:
Data directory.
"""
# open the destination file for writing
fd = open(datadir + '/post_evaluation.dat', 'w')
# write the header
fd.write('#--t-------------'.encode('utf-8'))
for variable in variables:
fd.write('--{0:15}'.format((variable+'--------------------')[:15]).encode('utf-8'))
fd.write('\n')
# read the list of all VAR files
var_list_file = open('data/proc0/varN.list')
var_list = var_list_file.readlines()
var_list_file.close()
if (len(variables[0]) == 1):
variables = [variables]
# check if bb needs to be computed
bb_flag = False
# array containing the variables which depend on bb
b_dep = ['ab_int', 'jb_int', 'b1m', 'b2m', 'bm2', 'abm', 'abrms', 'jbm', 'brms', 'gffz']
if (len(set(variables + b_dep)) < len(variables + b_dep)):
bb_flag = True
# check if jj needs to be computed
jj_flag = False
# array containing the variables which depend on jj
j_dep = ['jb_int', 'j2m', 'jm2', 'jbm', 'jrms']
if (len(set(variables + j_dep)) < len(variables + j_dep)):
jj_flag = True
for var_file in var_list:
# read the var file
var = pc.read_var(varfile = var_file[:-1], datadir = datadir, quiet = True)
# the output string which will be written in the destination file
out_string = '{0:1.9e} '.format(np.float64(var.t))
aa = var.aa[:,var.n1:var.n2,var.m1:var.m2,var.l1:var.l2]
if bb_flag:
bb = pc.curl(var.aa, var.dx, var.dy, var.dz, var.x, var.y, var.z)
bb = bb[:,var.n1:var.n2,var.m1:var.m2,var.l1:var.l2]
if jj_flag:
jj = pc.curl2(var.aa, var.dx, var.dy, var.dz, var.x, var.y, var.z)
jj = jj[:,var.n1:var.n2,var.m1:var.m2,var.l1:var.l2]
for variable in variables:
if variable == 'b2m':
b2m = (pc.dot2(bb)).mean()
out_string += '{0:1.9e} '.format(np.float64(b2m))
elif variable == 'j2m':
j2m = (pc.dot2(jj)).mean()
out_string += '{0:1.9e} '.format(np.float64(j2m))
break
elif variable == 'abm':
abm = (pc.dot(var.aa, bb)).mean()
out_string += '{0:1.9e} '.format(np.float64(abm))
# generalized flux function (see Yeates, Hornig 2011)
elif variable == 'gffz':
gffz = np.sum(pc.dot(aa, bb)*np.sqrt(pc.dot2(bb)) / \
((bb[2,:,:,:]) * np.mean(np.mean(np.sqrt(pc.dot2(bb)), axis = 2), axis = 1)))
out_string += '{0:1.9e} '.format(np.float64(gffz))
fd.write((out_string+'\n').encode('utf-8'))
fd.close()
def post_compute(variables = ['b2m'], datadir = 'data'):
"""
Compute diagnostic variables from the VAR files.
call signature::
post_compute(variables = ['urms'], datadir = 'data')
Read the VAR files and compute the diagnostic variables.
Write the result in 'post_evaluation.dat'.
Keyword arguments:
*variables*:
The diagnostic variables to be computed.
*datadir*:
Data directory.
"""
# open the destination file for writing
fd = open(datadir + '/post_evaluation.dat', 'w')
# write the header
fd.write('#--t-------------'.encode('utf-8'))
for variable in variables:
fd.write('--{0:15}'.format((variable+'--------------------')[:15]).encode('utf-8'))
fd.write('\n')
# read the list of all VAR files
var_list_file = open('data/proc0/varN.list')
var_list = var_list_file.readlines()
var_list_file.close()
if (len(variables[0]) == 1):
variables = [variables]
# check if bb needs to be computed
bb_flag = False
# array containing the variables which depend on bb
b_dep = ['ab_int', 'jb_int', 'b1m', 'b2m', 'bm2', 'abm', 'abrms', 'jbm', 'brms', 'gffz']
if (len(set(variables + b_dep)) < len(variables + b_dep)):
bb_flag = True
# check if jj needs to be computed
jj_flag = False
# array containing the variables which depend on jj
j_dep = ['jb_int', 'j2m', 'jm2', 'jbm', 'jrms']
if (len(set(variables + j_dep)) < len(variables + j_dep)):
jj_flag = True
for var_file in var_list:
# read the var file
var = pc.read_var(varfile = var_file[:-1], datadir = datadir, quiet = True)
# the output string which will be written in the destination file
out_string = '{0:1.9e} '.format(np.float64(var.t))
aa = var.aa[:,var.n1:var.n2,var.m1:var.m2,var.l1:var.l2]
if bb_flag:
bb = pc.curl(var.aa, var.dx, var.dy, var.dz)
bb = bb[:,var.n1:var.n2,var.m1:var.m2,var.l1:var.l2]
if jj_flag:
jj = pc.curl2(var.aa, var.dx, var.dy, var.dz)
jj = jj[:,var.n1:var.n2,var.m1:var.m2,var.l1:var.l2]
for variable in variables:
if variable == 'b2m':
b2m = (pc.dot2(bb)).mean()
out_string += '{0:1.9e} '.format(np.float64(b2m))
elif variable == 'j2m':
j2m = (pc.dot2(jj)).mean()
out_string += '{0:1.9e} '.format(np.float64(j2m))
break
elif variable == 'abm':
abm = (pc.dot(var.aa, bb)).mean()
out_string += '{0:1.9e} '.format(np.float64(abm))
# generalized flux function (see Yeates, Hornig 2011)
elif variable == 'gffz':
gffz = np.sum(pc.dot(aa, bb)*np.sqrt(pc.dot2(bb)) / \
((bb[2,:,:,:]) * np.mean(np.mean(np.sqrt(pc.dot2(bb)), axis = 2), axis = 1)))
out_string += '{0:1.9e} '.format(np.float64(gffz))
fd.write((out_string+'\n').encode('utf-8'))
fd.close()
