def new_parent_selection(config, system, q=False, verbose=False):
"""
takes the config file and selects new parents based on the criteria therein
returns [(O,mode), ...]
"""
# selection = dict( config.items("parents") )
selection = dict( [(key.lower(),value) for key, value in config.items("parents")] )
network = system.network
freqs = system.compute_3mode_freqs()
if verbose: print "selection : ", selection
### modeNos
if selection.has_key("modeNos".lower()):
if verbose: print "selecting parents based on modeNo"
modeNos = [int(l) for l in selection["modeNos".lower()].strip().split()]
else:
modeNos = range(len(network))
### genNos
if selection.has_key("genNos".lower()):
_modeNos = []
if verbose: print "selecting parents based on genNo"
gens,_ = network.gens()
for genNo in [int(l) for l in selection['genNos'.lower()].split()]:
for modeNo in gens[genNo]:
if modeNo in modeNos:
_modeNos.append( modeNo )
modeNos = _modeNos
### l
has_min = selection.has_key("min_l".lower())
has_max = selection.has_key("max_l".lower())
if has_min or has_max:
if verbose: print "selecting parents based on l"
_modeNos = []
for modeNo in modeNos:
l = network.modes[modeNo].l
if (not has_min) or (has_min and (l >= int(selection["min_l".lower()]))):
if (not has_max) or (has_max and (l <= int(selection["max_l".lower()]))):
_modeNos.append( modeNo )
modeNos = _modeNos
### frac_Oorb
has_min = selection.has_key("min_frac_Oorb".lower())
has_max = selection.has_key("max_frac_Oorb".lower())
if has_min or has_max:
if verbose: print "selecting parents based on frac_Oorb"
_modeNos = []
for modeNo in modeNos:
absO_Oorb = abs(freqs[modeNo]/system.Oorb)
if (not has_min) or (has_min and (absO_Oorb >= float(selection["min_frac_Oorb".lower()]))):
if (not has_max) or (has_max and (absO_Oorb <= float(selection["max_frac_Oorb".lower()]))):
_modeNos.append( modeNo )
modeNos = _modeNos
### Amean
if selection.has_key("min_Amean".lower()):
if verbose: print "selecting parents based on min_Amean"
keep, _ = pn.downselect_Amean(q, float(selection["min_Amean".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
if selection.has_key("max_Amean"):
if verbose: print "selecting parents based on max_Amean"
_, keep = pn.downselect_Amean(q, float(selection["max_Amean".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
### Alast
if selection.has_key("min_Alast".lower()):
if verbose: print "selecting parents based on min_Alast"
keep, _ = pn.downselect_Alast(q, float(selection["min_Alast".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
if selection.has_key("max_Alast"):
if verbose: print "selecting parents based on max_Alast"
_, keep = pn.downselect_Alast(q, float(selection["max_Alast".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
### Afit
if selection.has_key("min_Afit".lower()):
if verbose: print "selecting parents based on min_Afit"
keep, _ = pn.downselect_Afit(q, float(selection["min_Afit".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
if selection.has_key("max_Afit"):
if verbose: print "selecting parents based on max_Afit"
_, keep = pn.downselect_Afit(q, float(selection["max_Afit".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
### num_k
if selection.has_key("min_num_k".lower()):
if verbose: print "selecting parents based on min_num_k"
keep, _ = pn.downselect_num_k(q, float(selection["min_num_k".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
if selection.has_key("max_num_k"):
if verbose: print "selecting parents based on max_num_k"
_, keep = pn.downselect_num_k(q, float(selection["max_num_k".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
### min_heuristic
if selection.has_key("min_heuristic".lower()):
if verbose: print "selecting parents based on min_heuristic"
keep, _ = pn.downselect_min_heuristic(q, float(selection["min_heuristic".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
if selection.has_key("max_heuristic"):
if verbose: print "selecting parents based on max_heuristic"
_, keep = pn.downselect_min_heuristic(q, float(selection["max_heuristic".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
### min_Ethr
if selection.has_key("min_Ethr".lower()):
if verbose: print "selecting parents based on min_Ethr"
keep, _ = pn.downselect_min_Ethr(q, float(selection["min_Ethr".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
if selection.has_key("max_Ethr"):
if verbose: print "selecting parents based on max_Ethr"
_, keep = pn.downselect_min_Ethr(q, float(selection["max_Ethr".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
### collE
if selection.has_key("min_collE".lower()):
if verbose: print "selecting parents based on min_collE"
keep, _ = pn.downselect_collE(q, float(selection["min_collE".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
if selection.has_key("max_collE"):
if verbose: print "selecting parents based on max_collE"
_, keep = pn.downselect_collE(q, float(selection["max_collE".lower()]), network, mode_nums=modeNos)
modeNos = [network.modeNoD[nlms] for nlms in keep]
return [(freqs[modeNo], network.modes[modeNo]) for modeNo in sorted(modeNos)]
network = system.network
Oorb = system.Oorb
Porb = 2*np.pi/Oorb
if opts.verbose: print "reading time-domain data from %s" % opts.filename
t_P, q, N_m = nm_u.load_out(opts.filename, tmin=opts.min_t_Porb, tmax=opts.max_t_Porb, downsample=opts.file_downsample)
remove = []
for Athr, method in zip(Athrs, downselection_methods):
if opts.verbose: _athr, _athrexp = nm_u.float_to_scientific(Athr)
if method == "Alast":
if opts.verbose: print "downselecting modes using %s and Athr=%fe%d" % (method, _athr, _athrexp)
remove += pn.downselect_Alast(q, Athr, network)[1]
elif method == "Amean":
if opts.verbose: print "downselecting modes using %s and Athr=%fe%d" % (method, _athr, _athrexp)
remove += pn.downselect_Amean(q, Athr, network)[1]
elif method == "Afit":
if opts.verbose: print "downselecting modes using %s and Athr=%fe%d" % (method, _athr, _athrexp)
remove += pn.downselect_Afit(q, Athr, network, t_P=t_P)[1]
else:
raise ValueError, "unknown downselection method: %s\nplease supply at last one downselection method from the following:\n\tAlst\n\tAmean\n\tAfit"
remove = list(set(remove)) # generate a unique set of modes to remove
if opts.verbose: print "attempting to remove %d/%d modes from the network" % (len(remove), N_m)
network = network.remove_modes(remove, verbose=opts.verbose)
if opts.verbose: print "writing network to %s" % opts.new_logfilename
nm_u.write_log(opts.new_logfilename, system)