def extract_chain_single(name,
sim_name,
start_idx=1,
zlens=None,
sigmasubmax=None,
observed_fluxes=None,
mhalomin=None,
mhmmax=None):
chain_info_path = chainpath_out + 'raw_chains_mcrelation/' + name + '/simulation_info.txt'
params_varied, varyparams_info, Ncores, cores_per_lens = read_chain_info(
chain_info_path)
#lens_config, lens_R_index = read_R_index(chainpath_out+chain_name+'/R_index_config.txt',0)
chain_file_path = chainpath_out + 'raw_chains_mcrelation/' + name + '/chain'
params_header = None
order = None
print(observed_fluxes)
if ~os.path.exists(chainpath_out + 'processed_chains/' + sim_name + '/'):
create_directory(chainpath_out + 'processed_chains/' + sim_name + '/')
if ~os.path.exists(chainpath_out + 'processed_chains/' + sim_name + '/' +
name):
create_directory(chainpath_out + 'processed_chains/' + sim_name + '/' +
name)
copy_directory(
chain_info_path,
chainpath_out + 'processed_chains/' + sim_name + '/' + name + '/')
end = int(start_idx + cores_per_lens)
init = True
for i in range(start_idx + 1, end):
folder_name = chain_file_path + str(i) + '/'
#print(folder_name)
try:
fluxes = np.loadtxt(folder_name + '/fluxes.txt')
if observed_fluxes is None:
obs_data = read_data(folder_name + '/lensdata.txt')
observed_fluxes = obs_data[0].m
params = np.loadtxt(folder_name + '/parameters.txt', skiprows=1)
zd, sat = None, None
if 'lens_redshift' in params_varied:
col = params_varied.index('lens_redshift')
zd = params[:, col]
params = np.delete(params, col, 1)
else:
assert zlens is not None
zd = np.array([zlens] * np.shape(params)[0])
if 'satellite_thetaE' in params_varied:
col = params_varied.index('satellite_thetaE')
sat = params[:, col:(col + 3)]
if 'satellite_thetaE_1' in params_varied:
col = params_varied.index('satellite_thetaE_1')
sat = params[:, col:(col + 6)]
params = params[:, 0:9]
# SIE_gamma
params = np.delete(params, 1, 1)
macro_model = np.loadtxt(folder_name + '/macro.txt')
macro_model[:, 3] = transform_ellip(macro_model[:, 3])
if zd is not None:
macro_model = np.column_stack((macro_model, zd))
if sat is not None:
macro_model = np.column_stack((macro_model, sat))
if sigmasubmax is not None:
keep = np.where(params[:, 1] < sigmasubmax)[0]
else:
keep = np.where(params[:, 1] < 10000)[0]
fluxes = fluxes[keep, :]
params = params[keep, :]
macro_model = macro_model[keep, :]
if mhalomin is not None:
keep = np.where(params[:, 3] > mhalomin)[0]
else:
keep = np.where(params[:, 3] < 10000)[0]
fluxes = fluxes[keep, :]
params = params[keep, :]
macro_model = macro_model[keep, :]
if mhmmax is not None:
keep = np.where(params[:, 5] < mhmmax)[0]
else:
keep = np.where(params[:, 5] < 10000)[0]
fluxes = fluxes[keep, :]
params = params[keep, :]
macro_model = macro_model[keep, :]
assert fluxes.shape[0] == params.shape[0]
except:
#print('didnt find a file... '+str(chain_file_path + str(i)+'/'))
continue
if params_header is None:
with open(folder_name + '/parameters.txt', 'r') as f:
lines = f.read().splitlines()
head = lines[0].split(' ')
params_header = ''
for word in head:
if word not in ['#', '']:
params_header += word + ' '
if init:
lens_fluxes = fluxes
lens_params = params
lens_all = np.column_stack((params, macro_model))
init = False
else:
lens_fluxes = np.vstack((lens_fluxes, fluxes))
lens_params = np.vstack((lens_params, params))
new = np.column_stack((params, macro_model))
lens_all = np.vstack((lens_all, new))
return lens_fluxes, observed_fluxes.reshape(
1, 4), lens_params, lens_all, params_header
def process_raw(name,
Npert,
sim_name='grism_quads',
keep_N=3000,
sigmasubmax=None,
mhalomin=None,
mhmmax=None,
deplete=False):
"""
coverts output from cluster into single files for each lens
"""
header = 'f1 f2 f3 f4 stat srcsize sigmasub deltalos mparent alpha c0 beta zeta re gx gy eps epstheta shear sheartheta gmacro'
header += ' lens_redshift'
keep_inds = [0, 1, 2, 3]
uncertainty_in_ratios = False
zlens = None
observed_fluxes = None
run_name = deepcopy(name)
keep_p = True
samples_completed = 500000
if name[0:8] == 'lens1422':
run_name = ['lens1422_mcrelation']
run_name += [run_name[0] + '_extended']
zlens = 0.36
sigmas = [0.011, 0.01, 0.013]
keep_inds = [0, 1, 2]
elif name[0:8] == 'lens0435':
run_name = ['lens0435_mcrelation']
run_name += [run_name[0] + '_extended']
zlens = 0.45
header += ' satellite_thetaE satellite_x satellite_y'
sigmas = [0.05, 0.049, 0.048, 0.056]
elif name[0:8] == 'lens2038':
run_name = ['lens2038_mcrelation']
run_name += [run_name[0] + '_extended']
keep_p = False
zlens = 0.23
sigmas = [0.01, 0.017, 0.022, 0.022]
elif name[0:8] == 'lens1606':
run_name = ['lens1606_mcrelation']
run_name += [run_name[0] + '_extended']
header += ' satellite_thetaE satellite_x satellite_y'
sigmas = [0.03, 0.03, 0.02 / 0.59, 0.02 / 0.79]
#sigmas = [0.00001]*4
elif name[0:8] == 'lens0405':
run_name = ['lens0405_mcrelation']
run_name += [run_name[0] + '_extended']
run_name += [run_name[0] + '_supplement']
#sigmas = [0.04, 0.03/0.7, 0.04/1.28, 0.05/0.94]
observed_fluxes = np.array([1., 0.65, 1.25, 1.17]) * 1.25**-1
sigmas = [
0.04 * observed_fluxes[0]**-1, 0.04 * observed_fluxes[1]**-1,
0.03**observed_fluxes[2]**-1, 0.04 * observed_fluxes[3]**-1
]
elif name[0:8] == 'lens2033':
run_name = ['lens2033_mcrelation']
run_name += [run_name[0] + '_extended']
header += ' satellite_thetaE_1 satellite_x_1 satellite_y_1 satellite_thetaE_2 satellite_x_2 satellite_y_2'
zlens = 0.66
sigmas = np.array([0.03, 0.03 / 0.65, 0.02 / 0.5, 0.02 / 0.53])
#sigmas = np.array([0.2, 0.2/0.65, 0.2/0.5, 0.2/0.53])
elif name[0:8] == 'lens2026':
run_name = ['lens2026_mcrelation']
run_name += [run_name[0] + '_extended']
sigmas = [0.02, 0.02 / 0.75, 0.01 / 0.31, 0.01 / 0.28]
elif name[0:8] == 'lens0911':
keep_p = False
#run_name = ['lens0911_varyshear_extended',]
#run_name += ['lens0911_old', 'lens0911_aurora']
run_name = ['lens0911_mcrelation', 'lens0911_mcrelation_benson']
#run_name = ['lens0911_varyshear_hoffman_2']
header += ' satellite_thetaE satellite_x satellite_y'
zlens = 0.77
sigmas = [0.04 / 0.56, 0.05, 0.04 / 0.53, 0.04 / 0.24]
elif name[0:8] == 'lens0128':
run_name = ['lens0128_mcrelation']
run_name += [run_name[0] + '_extended']
zlens = 1.145
uncertainty_in_ratios = True
keep_inds = [0, 1, 2]
sigmas = [0.029, 0.029, 0.032]
elif name[0:8] == 'lens1115':
run_name = ['lens1115_mcrelation']
run_name += [run_name[0] + '_extended']
zlens = 0.31
uncertainty_in_ratios = True
keep_inds = [0, 1, 2]
#sigmas = [0.1, 0.1, 0.1]
sigmas = [0.1] * 3
elif name[0:8] == 'lens0414':
run_name = ['lens0414_mcrelation']
run_name += [run_name[0] + '_extended']
#observed_fluxes = np.array([1, 0.83, 0.36, 0.16])
#sigmas = [0.1, 0.1, 0.1]
sigmas = [0.05] * 3
observed_fluxes = np.array([1, 0.903, 0.389, 0.145])
#sigmas = [0.05, 0.04, 0.04]
zlens = 0.96
uncertainty_in_ratios = True
keep_inds = [0, 1, 2]
sigmas = np.array(sigmas)
print('loading chains... ')
fluxes, fluxes_obs, parameters, all, _ = extract_chain(
run_name,
sim_name,
zlens=zlens,
sigmasubmax=sigmasubmax,
observed_fluxes=observed_fluxes,
mhalomin=mhalomin,
keep_p=keep_p,
mhmmax=mhmmax)
print('done.')
all = np.squeeze(all)
fluxes, fluxes_obs = np.squeeze(fluxes), np.squeeze(fluxes_obs)
if mhmmax is not None:
chain_file_path = chainpath_out + 'processed_chains/grism_quads_mcrelation_CDM/' + name + '/'
elif deplete is True:
#print(deplete is True)
#fluxes = fluxes[0::2, :]
fluxes = np.delete(fluxes, np.arange(0, fluxes.shape[0], 3), axis=0)
all = np.delete(all, np.arange(0, all.shape[0], 3), axis=0)
#all = all[0::2, :]
chain_file_path = chainpath_out + 'processed_chains/grism_quads_depleted_mcrelation/' + name + '/'
else:
chain_file_path = chainpath_out + 'processed_chains/grism_quads_mcrelation/' + name + '/'
print('nrealizations: ', fluxes.shape[0])
if ~os.path.exists(chain_file_path):
create_directory(chain_file_path)
print('sampling flux uncertainties, lens ' + name[0:8] + '... ')
inds_to_keep_list, statistics = add_flux_perturbations(
fluxes, fluxes_obs, sigmas, Npert, keep_inds, uncertainty_in_ratios)
print('done.')
for i, indexes in enumerate(inds_to_keep_list):
f = fluxes[indexes[0:keep_N], :]
final_fluxes = np.column_stack((f, np.array(statistics[i][0:keep_N])))
x = np.column_stack((final_fluxes, all[indexes[0:keep_N], :]))
np.savetxt(chain_file_path + 'samples' + str(i + 1) + '.txt',
x,
fmt='%.5f',
header=header)
def process_raw(name,
Npert,
sim_name='grism_quads',
keep_N=3000,
sigmasubmax=None,
mhalomin=None,
mhmmax=None,
deplete=False):
"""
coverts output from cluster into single files for each lens
"""
header = 'f1 f2 f3 f4 stat srcsize sigmasub deltalos mparent alpha mhm re gx gy eps epstheta shear sheartheta gmacro'
header += ' re_normed gx_normed gy_normed eps_normed epstheta_normed shear_normed sheartheta_normed gmacro_normed'
header += ' lens_redshift'
keep_inds = [0, 1, 2, 3]
uncertainty_in_ratios = False
zlens = None
observed_fluxes = None
run_name = deepcopy(name)
keep_p = True
samples_completed = 500000
if name[0:8] == 'lens1422':
run_name = ['lens1422_CDM_test']
mhm_flags = [True] + [False] * (len(run_name) - 1)
zlens = 0.36
sigmas = [0.011, 0.01, 0.013]
keep_inds = [0, 1, 2]
sigmas = np.array(sigmas)
if mhm_flags is None:
mhm_flags = [False] * len(run_name)
else:
print(mhm_flags, run_name)
assert len(mhm_flags) == len(run_name)
print('loading chains... ')
fluxes, fluxes_obs, parameters, all, _ = extract_chain(
run_name,
sim_name,
zlens=zlens,
sigmasubmax=sigmasubmax,
observed_fluxes=observed_fluxes,
mhalomin=mhalomin,
keep_p=keep_p,
mhmmax=mhmmax,
mhm_flags=mhm_flags)
print('done.')
all = np.squeeze(all)
fluxes, fluxes_obs = np.squeeze(fluxes), np.squeeze(fluxes_obs)
chain_file_path = chainpath_out + 'processed_chains/grism_quads_CDM/lens1422_CDM_test/'
print('nrealizations: ', fluxes.shape[0])
if ~os.path.exists(chain_file_path):
create_directory(chain_file_path)
print('sampling flux uncertainties, lens ' + name[0:8] + '... ')
inds_to_keep_list, statistics = add_flux_perturbations(
fluxes, fluxes_obs, sigmas, Npert, keep_inds, uncertainty_in_ratios)
print('done.')
for i, indexes in enumerate(inds_to_keep_list):
f = fluxes[indexes[0:keep_N], :]
final_fluxes = np.column_stack((f, np.array(statistics[i][0:keep_N])))
x = np.column_stack((final_fluxes, all[indexes[0:keep_N], :]))
np.savetxt(chain_file_path + 'samples' + str(i + 1) + '.txt',
x,
fmt='%.5f',
header=header)