def piE_tE_deltac():
span = 0.999999426697
smooth = 0.02
quantiles_2d = None
hist2d_kwargs = None
labels = None
label_kwargs = None
show_titles = False
title_fmt = ".2f"
title_kwargs = None
"""
!!! NOTE: CHOICE OF THE quantiles_2d HAS A LARGE EFFECT
ON THE WAY THIS PLOT LOOKS !!!
Plot piE-tE 2D posteriors from OGLE photometry only fits.
Also plot PopSyCLE simulations simultaneously.
"""
# Initialize values.
if label_kwargs is None:
label_kwargs = dict()
if title_kwargs is None:
title_kwargs = dict()
if hist2d_kwargs is None:
hist2d_kwargs = dict()
# Set defaults.
hist2d_kwargs['alpha'] = hist2d_kwargs.get('alpha', 0.2)
hist2d_kwargs['levels'] = hist2d_kwargs.get('levels', quantiles_2d)
ob120169_yaml = open(ob120169_dir + ob120169_id + '_params.yaml').read()
params_120169 = yaml.safe_load(ob120169_yaml)
ob140613_yaml = open(ob140613_dir + ob140613_id + '_params.yaml').read()
params_140613 = yaml.safe_load(ob140613_yaml)
ob150029_yaml = open(ob150029_dir + ob150029_id + '_params.yaml').read()
params_150029 = yaml.safe_load(ob150029_yaml)
ob150211_yaml = open(ob150211_dir + ob150211_id + '_params.yaml').read()
params_150211 = yaml.safe_load(ob150211_yaml)
# OB120169 fit results.
data_120169 = munge.getdata2('ob120169',
phot_data=params_120169['phot_data'],
ast_data=params_120169['astrom_data'])
fitter_120169 = model_fitter.PSPL_Solver(
data_120169,
getattr(model, params_120169['model']),
add_error_on_photometry=params_120169['add_error_on_photometry'],
multiply_error_on_photometry=params_120169[
'multiply_error_on_photometry'],
outputfiles_basename=ob120169_dir + ob120169_id + '_')
results_120169 = fitter_120169.load_mnest_results_for_dynesty()
smy_120169 = fitter_120169.load_mnest_summary()
# OB140613 fit results.
data_140613 = munge.getdata2('ob140613',
phot_data=params_140613['phot_data'],
ast_data=params_140613['astrom_data'])
fitter_140613 = model_fitter.PSPL_Solver(
data_140613,
getattr(model, params_140613['model']),
add_error_on_photometry=params_140613['add_error_on_photometry'],
multiply_error_on_photometry=params_140613[
'multiply_error_on_photometry'],
outputfiles_basename=ob140613_dir + ob140613_id + '_')
results_140613 = fitter_140613.load_mnest_results_for_dynesty()
smy_140613 = fitter_140613.load_mnest_summary()
# OB150029 fit results.
data_150029 = munge.getdata2('ob150029',
phot_data=params_150029['phot_data'],
ast_data=params_150029['astrom_data'])
fitter_150029 = model_fitter.PSPL_Solver(
data_150029,
getattr(model, params_150029['model']),
add_error_on_photometry=params_150029['add_error_on_photometry'],
multiply_error_on_photometry=params_150029[
'multiply_error_on_photometry'],
outputfiles_basename=ob150029_dir + ob150029_id + '_')
results_150029 = fitter_150029.load_mnest_results_for_dynesty()
smy_150029 = fitter_150029.load_mnest_summary()
# OB150211 fit results.
data_150211 = munge.getdata2('ob150211',
phot_data=params_150211['phot_data'],
ast_data=params_150211['astrom_data'])
fitter_150211 = model_fitter.PSPL_Solver(
data_150211,
getattr(model, params_150211['model']),
add_error_on_photometry=params_150211['add_error_on_photometry'],
multiply_error_on_photometry=params_150211[
'multiply_error_on_photometry'],
outputfiles_basename=ob150211_dir + ob150211_id + '_')
results_150211 = fitter_150211.load_mnest_results_for_dynesty()
smy_150211 = fitter_150211.load_mnest_summary()
# Extract weighted samples.
samples_120169 = results_120169['samples']
samples_140613 = results_140613['samples']
samples_150029 = results_150029['samples']
samples_150211 = results_150211['samples']
try:
weights_120169 = np.exp(results_120169['logwt'] -
results_120169['logz'][-1])
weights_140613 = np.exp(results_140613['logwt'] -
results_140613['logz'][-1])
weights_150029 = np.exp(results_150029['logwt'] -
results_150029['logz'][-1])
weights_150211 = np.exp(results_150211['logwt'] -
results_150211['logz'][-1])
except:
weights_120169 = results_120169['weights']
weights_140613 = results_140613['weights']
weights_150029 = results_150029['weights']
weights_150211 = results_150211['weights']
# Deal with 1D results. A number of extra catches are also here
# in case users are trying to plot other results besides the `Results`
# instance generated by `dynesty`.
samples_120169 = np.atleast_1d(samples_120169)
if len(samples_120169.shape) == 1:
samples_120169 = np.atleast_2d(samples_120169)
else:
assert len(samples_120169.shape) == 2, "Samples must be 1- or 2-D."
samples_120169 = samples_120169.T
assert samples_120169.shape[0] <= samples_120169.shape[1], "There are more " \
"dimensions than samples!"
samples_140613 = np.atleast_1d(samples_140613)
if len(samples_140613.shape) == 1:
samples_140613 = np.atleast_2d(samples_140613)
else:
assert len(samples_140613.shape) == 2, "Samples must be 1- or 2-D."
samples_140613 = samples_140613.T
assert samples_140613.shape[0] <= samples_140613.shape[1], "There are more " \
"dimensions than samples!"
samples_150029 = np.atleast_1d(samples_150029)
if len(samples_150029.shape) == 1:
samples_150029 = np.atleast_2d(samples_150029)
else:
assert len(samples_150029.shape) == 2, "Samples must be 1- or 2-D."
samples_150029 = samples_150029.T
assert samples_150029.shape[0] <= samples_150029.shape[1], "There are more " \
"dimensions than samples!"
samples_150211 = np.atleast_1d(samples_150211)
if len(samples_150211.shape) == 1:
samples_150211 = np.atleast_2d(samples_150211)
else:
assert len(samples_150211.shape) == 2, "Samples must be 1- or 2-D."
samples_150211 = samples_150211.T
assert samples_150211.shape[0] <= samples_150211.shape[1], "There are more " \
"dimensions than samples!"
# Maximum likelihood vals
smy_list = [smy_120169, smy_140613, smy_150029, smy_150211]
smy_name = ['OB120169', 'OB140613', 'OB150029', 'OB150211']
maxl = {}
for ss, smy in enumerate(smy_list):
print(smy_name[ss])
print('tE : ', smy['MaxLike_tE'][0])
print('piE : ',
np.hypot(smy['MaxLike_piE_E'][0], smy['MaxLike_piE_N'][0]))
maxl[smy_name[ss]] = {
'tE': smy['MaxLike_tE'][0],
'piE': np.hypot(smy['MaxLike_piE_E'][0], smy['MaxLike_piE_N'][0])
}
#############
# The observations.
# OB110022 from Lu+16.
# Finished targets are phot + astrom solutions
# Ongoing targets are phot parallax solutions
# (global MEDIAN +/- 1 sigma)
#############
# OB110022
piEE_110022 = -0.393
piEE_110022_pe = 0.013
piEE_110022_me = 0.012
piEN_110022 = -0.071
piEN_110022_pe = 0.013
piEN_110022_me = 0.012
piE_110022, piE_110022_pe, piE_110022_me = calc_hypot_and_err(
piEE_110022, piEE_110022_pe, piEE_110022_me, piEN_110022,
piEN_110022_pe, piEN_110022_me)
tE_110022 = 61.4
tE_110022_pe = 1.0
tE_110022_me = 1.0
# This is an upper limit.
dcmax_110022 = 2.19 / np.sqrt(8)
dcmax_110022_pe = 1.06 / np.sqrt(8)
dcmax_110022_me = 1.17 / np.sqrt(8)
# Plot the piE-tE 2D posteriors.
# tE = 2; piEE,N = 3, 4
fig, ax = plt.subplots(1,
2,
figsize=(14, 6),
sharey=True,
gridspec_kw={'width_ratios': [1, 1.4]})
plt.subplots_adjust(left=0.1, bottom=0.15, wspace=0.1)
tE_120169 = samples_120169[2]
tE_140613 = samples_140613[2]
tE_150029 = samples_150029[2]
tE_150211 = samples_150211[2]
thetaE_120169 = samples_120169[3]
thetaE_140613 = samples_140613[3]
thetaE_150029 = samples_150029[3]
thetaE_150211 = samples_150211[3]
piE_120169 = np.hypot(samples_120169[5], samples_120169[6])
piE_140613 = np.hypot(samples_140613[5], samples_140613[6])
piE_150029 = np.hypot(samples_150029[5], samples_150029[6])
piE_150211 = np.hypot(samples_150211[5], samples_150211[6])
sx = smooth
sy = smooth
hist2d_kwargs['fill_contours'] = hist2d_kwargs.get('fill_contours', False)
hist2d_kwargs['plot_contours'] = hist2d_kwargs.get('plot_contours', True)
ax[0].errorbar(dcmax_110022,
piE_110022,
xerr=np.array([[dcmax_110022_me], [dcmax_110022_pe]]),
fmt='*',
color='cyan',
markersize=15,
xuplims=True)
model_fitter.contour2d_alpha(thetaE_120169 / np.sqrt(8),
piE_120169,
span=[span, span],
quantiles_2d=quantiles_2d,
weights=weights_120169,
ax=ax[0],
smooth=[sy, sx],
color='blue',
**hist2d_kwargs,
plot_density=False)
model_fitter.contour2d_alpha(thetaE_140613 / np.sqrt(8),
piE_140613,
span=[span, span],
quantiles_2d=quantiles_2d,
weights=weights_140613,
ax=ax[0],
smooth=[sy, sx],
color='hotpink',
**hist2d_kwargs,
plot_density=False)
model_fitter.contour2d_alpha(thetaE_150029 / np.sqrt(8),
piE_150029,
span=[span, span],
quantiles_2d=quantiles_2d,
weights=weights_150029,
ax=ax[0],
smooth=[sy, sx],
color='red',
**hist2d_kwargs,
plot_density=False)
model_fitter.contour2d_alpha(thetaE_150211 / np.sqrt(8),
piE_150211,
span=[span, span],
quantiles_2d=quantiles_2d,
weights=weights_150211,
ax=ax[0],
smooth=[sy, sx],
color='dodgerblue',
**hist2d_kwargs,
plot_density=False)
ax[1].plot(tE_110022, piE_110022, color='cyan', marker='*', ms=15)
model_fitter.contour2d_alpha(tE_120169,
piE_120169,
span=[span, span],
quantiles_2d=quantiles_2d,
weights=weights_120169,
ax=ax[1],
smooth=[sy, sx],
color='blue',
**hist2d_kwargs,
plot_density=False)
model_fitter.contour2d_alpha(tE_150211,
piE_150211,
span=[span, span],
quantiles_2d=quantiles_2d,
weights=weights_150211,
ax=ax[1],
smooth=[sy, sx],
color='dodgerblue',
**hist2d_kwargs,
plot_density=False)
ax[1].plot(maxl['OB140613']['tE'],
maxl['OB140613']['piE'],
color='hotpink',
marker='*',
ms=15)
ax[1].plot(maxl['OB150029']['tE'],
maxl['OB150029']['piE'],
color='red',
marker='*',
ms=15)
ax[1].plot(0.01, 100, color='cyan', label='OB110022')
ax[1].plot(0.01, 100, color='blue', label='OB120169')
ax[1].plot(0.01, 100, color='hotpink', label='OB140613')
ax[1].plot(0.01, 100, color='red', label='OB150029')
ax[1].plot(0.01, 100, color='dodgerblue', label='OB150211')
# Add the PopSyCLE simulation points.
# NEED TO UPDATE THIS WITH BUGFIX IN DELTAM
t = Table.read(
'/u/casey/scratch/papers/microlens_2019/popsycle_rr_files/Mock_EWS_v2.fits'
)
bh_idx = np.where(t['rem_id_L'] == 103)[0]
ns_idx = np.where(t['rem_id_L'] == 102)[0]
wd_idx = np.where(t['rem_id_L'] == 101)[0]
st_idx = np.where(t['rem_id_L'] == 0)[0]
u0_arr = t['u0']
thetaE_arr = t['theta_E']
# Stores the maximum astrometric shift
final_delta_arr = np.zeros(len(u0_arr))
# Stores the lens-source separation corresponding
# to the maximum astrometric shift
final_u_arr = np.zeros(len(u0_arr))
# Sort by whether the maximum astrometric shift happens
# before or after the maximum photometric amplification
big_idx = np.where(u0_arr > np.sqrt(2))[0]
small_idx = np.where(u0_arr <= np.sqrt(2))[0]
# Flux ratio of lens to source (and make it 0 if dark lens)
g_arr = 10**(-0.4 * (t['ubv_i_app_L'] - t['ubv_i_app_S']))
g_arr = np.nan_to_num(g_arr)
for i in np.arange(len(u0_arr)):
g = g_arr[i]
thetaE = thetaE_arr[i]
# Try all values between u0 and sqrt(2) to find max
# astrometric shift
if u0_arr[i] < np.sqrt(2):
u_arr = np.linspace(u0_arr[i], np.sqrt(2), 100)
delta_arr = np.zeros(len(u_arr))
for j in np.arange(len(u_arr)):
u = u_arr[j]
numer = 1 + g * (u**2 - u * np.sqrt(u**2 + 4) + 3)
denom = u**2 + 2 + g * u * np.sqrt(u**2 + 4)
delta = (u * thetaE / (1 + g)) * (numer / denom)
delta_arr[j] = delta
max_idx = np.argmax(delta_arr)
final_delta_arr[i] = delta_arr[max_idx]
final_u_arr[i] = u_arr[max_idx]
# Maximum astrometric shift will occur at sqrt(2)
if u0_arr[i] > np.sqrt(2):
u = u0_arr[i]
numer = 1 + g * (u**2 - u * np.sqrt(u**2 + 4) + 3)
denom = u**2 + 2 + g * u * np.sqrt(u**2 + 4)
delta = (u * thetaE / (1 + g)) * (numer / denom)
final_delta_arr[i] = delta
final_u_arr[i] = u
ax[0].scatter(final_delta_arr[st_idx],
t['pi_E'][st_idx],
alpha=0.4,
marker='.',
s=25,
c='gold')
ax[0].scatter(final_delta_arr[wd_idx],
t['pi_E'][wd_idx],
alpha=0.4,
marker='.',
s=25,
c='goldenrod')
ax[0].scatter(final_delta_arr[ns_idx],
t['pi_E'][ns_idx],
alpha=0.4,
marker='.',
s=25,
c='sienna')
ax[0].scatter(final_delta_arr[bh_idx],
t['pi_E'][bh_idx],
alpha=0.8,
marker='.',
s=25,
c='black')
ax[1].scatter(t['t_E'][st_idx],
t['pi_E'][st_idx],
alpha=0.4,
marker='.',
s=25,
color='gold')
ax[1].scatter(t['t_E'][wd_idx],
t['pi_E'][wd_idx],
alpha=0.4,
marker='.',
s=25,
color='goldenrod')
ax[1].scatter(t['t_E'][ns_idx],
t['pi_E'][ns_idx],
alpha=0.4,
marker='.',
s=25,
color='sienna')
ax[1].scatter(t['t_E'][bh_idx],
t['pi_E'][bh_idx],
alpha=0.8,
marker='.',
s=25,
color='black')
# Trickery to make the legend darker
ax[1].scatter(0.01,
100,
alpha=0.8,
marker='o',
s=25,
label='Star',
color='gold')
ax[1].scatter(0.01,
100,
alpha=0.8,
marker='o',
s=25,
label='WD',
color='goldenrod')
ax[1].scatter(0.01,
100,
alpha=0.8,
marker='o',
s=25,
label='NS',
color='sienna')
ax[1].scatter(0.01,
100,
alpha=0.8,
marker='o',
s=25,
label='BH',
color='black')
ax[0].set_xlabel('$\delta_{c,max}$ (mas)')
ax[0].set_ylabel('$\pi_E$')
ax[0].set_xscale('log')
ax[0].set_yscale('log')
ax[0].set_xlim(0.005, 4)
ax[0].set_ylim(0.009, 0.5)
ax[1].set_xlim(10, 400)
ax[1].set_ylim(0.009, 0.5)
ax[1].set_xlabel('$t_E$ (days)')
ax[1].set_xscale('log')
ax[1].set_yscale('log')
box = ax[1].get_position()
ax[1].set_position([box.x0, box.y0, box.width * 0.7, box.height])
ax[1].legend(bbox_to_anchor=(1.5, 0.5), loc="center right")
plt.savefig('piE_tE_deltac.png')
results_mass_120169 = {}
results_mass_120169['weights'] = results_120169['weights']
results_mass_120169['logvol'] = results_120169['logvol']
results_mass_120169['samples'] = results_120169['samples'][:, 17].reshape(
len(results_120169['samples']), 1)
def fittable(align, parallax=False):
'''
Create a LaTeX table comparing photometric and astrometric models
for an alignment, with or without parallax.
'''
from microlens.jlu.targets import ob150211_model as ob_mod
data = ob_mod.getdata(align + 'points_d/')
if parallax:
par = '_par'
fit_ast = model_fitter.PSPL_parallax_Solver(data)
fit_phot = model_fitter.PSPL_phot_parallax_Solver(data)
else:
par = ''
fit_ast = model_fitter.PSPL_Solver(data)
fit_phot = model_fitter.PSPL_phot_Solver(data)
mod_ast = align + 'mnest_pspl' + par + '/'
mod_phot = align + 'mnest_pspl' + par + '_phot/'
fit_ast.outputfiles_basename = mod_ast + 'aa_'
fit_phot.outputfiles_basename = mod_phot + 'aa_'
# Setup
pars_list = [
'mL', 't0', 'tE', 'beta', 'dL', 'dS', 'muRel_E', 'muRel_N', 'thetaE',
'piE_E', 'piE_N', 'mag_base', 'blen_frac'
]
pars = {
'mL': [r'$M$ (M$_{\odot}$)', '${0:.2f}^{{+{1:.2f}}}_{{-{2:.2f}}}$'],
't0': ['$t_0$ (MJD)', '${0:.2f}^{{+{1:.2f}}}_{{-{2:.2f}}}$'],
'tE': ['$t_E$ (days)', '${0:.2f}^{{+{1:.2f}}}_{{-{2:.2f}}}$'],
'beta': ['$u_0$', '${0:.2f}^{{+{1:.2f}}}_{{-{2:.2f}}}$'],
'dL': ['$d_L$ (kpc)', '${0:.2f}^{{+{1:.2f}}}_{{-{2:.2f}}}$'],
'dS': ['$d_S$ (kpc)', '${0:.2f}^{{+{1:.2f}}}_{{-{2:.2f}}}$'],
'muRel_E': [
r'$\mu_{\text{rel}, E}$ (mas/yr)',
'${0:.2f}^{{+{1:.2f}}}_{{-{2:.2f}}}$'
],
'muRel_N': [
r'$\mu_{\text{rel}, N}$ (mas/yr)',
'${0:.2f}^{{+{1:.2f}}}_{{-{2:.2f}}}$'
],
'thetaE': [r'$\theta_E$ (mas)', '${0:.2f}^{{+{1:.2f}}}_{{-{2:.2f}}}$'],
'piE_E': [r'$\pi_{E,E}$', '${0:.3f}^{{+{1:.3f}}}_{{-{2:.3f}}}$'],
'piE_N': [r'$\pi_{E,N}$', '${0:.3f}^{{+{1:.3f}}}_{{-{2:.3f}}}$'],
'mag_base': ['$I_0$', '${0:.3f}^{{+{1:.3f}}}_{{-{2:.3f}}}$'],
'blen_frac': ['$f_b$', '${0:.2f}^{{+{1:.2f}}}_{{-{2:.2f}}}$']
}
# Get quantiles and fits
fit_pars_ast, q_ast = ob_mod.quantiles(fit_ast)
best_ast = fit_ast.get_best_fit()
fit_pars_phot, q_phot = ob_mod.quantiles(fit_phot)
best_phot = fit_phot.get_best_fit()
out = open(mod_ast + 'table.txt', 'w')
for pp in pars_list:
p = pars[pp]
if pp in fit_pars_ast:
w_ast = p[1].format(best_ast[pp], q_ast[pp][1], q_ast[pp][2])
else:
w_ast = '--'
if pp in fit_pars_phot:
w_phot = p[1].format(best_phot[pp], q_phot[pp][1], q_phot[pp][2])
else:
w_phot = '--'
out.write(p[0] + ' & ' + w_phot + ' & ' + w_ast + ' \\\\\n')
out.close()
print('Output: ' + mod_ast + 'table.txt')
def plot_lightcurves():
"""
Plot the lightcurves for MB09260 and MB10364.
"""
mb09260_data = munge.getdata2('mb09260', phot_data=['MOA', 'HST'], ast_data = [])
mb10364_data = munge.getdata2('mb10364', phot_data=['MOA', 'HST'], ast_data = [])
mb09260_dir = '/u/jlu/work/microlens/MB09260/a_2020_02_19/model_fits/202_fit_multiphot_parallax/a0_'
mb10364_dir = '/u/jlu/work/microlens/MB10364/a_2020_02_18/model_fits/202_fit_multiphot_parallax/a0_'
mb09260_fitter = model_fitter.PSPL_Solver(mb09260_data,
model.PSPL_Phot_Par_Param1,
importance_nested_sampling = False,
n_live_points = 400,
evidence_tolerance = 0.5,
sampling_efficiency = 0.8,
outputfiles_basename=mb09260_dir)
mb10364_fitter = model_fitter.PSPL_Solver(mb10364_data,
model.PSPL_Phot_Par_Param1,
importance_nested_sampling = False,
n_live_points = 400,
evidence_tolerance = 0.5,
sampling_efficiency = 0.8,
outputfiles_basename=mb10364_dir)
mb09260_mod = mb09260_fitter.get_best_fit_model(def_best='maxl')
mb10364_mod = mb10364_fitter.get_best_fit_model(def_best='maxl')
times_mod = np.arange(54000, 57000, 1)
fig = plt.figure(1, figsize=(14,6))
plt.clf()
plt.subplots_adjust(left=0.07, right=0.99, wspace=0.15)
ax1 = plt.subplot(1, 2, 1)
ax2 = plt.subplot(1, 2, 2)
mb09260_mod_mag1 = mb09260_mod.get_photometry(times_mod, 0)
mb09260_mod_mag2 = mb09260_mod.get_photometry(times_mod, 1)
mb10364_mod_mag1 = mb10364_mod.get_photometry(times_mod, 0)
mb10364_mod_mag2 = mb10364_mod.get_photometry(times_mod, 1)
# Rescaled MOA to match what's on Artemis
mb09260_mag1_rs = mb09260_data['mag1'] + 3.25
mb10364_mag1_rs = mb10364_data['mag1'] + 1.75
mb09260_mag2_rs = mb09260_data['mag2'] - 1.05087
mb10364_mag2_rs = mb10364_data['mag2'] - 1.8983
mb09260_mod_mag1 += 3.25
mb10364_mod_mag1 += 1.75
mb09260_mod_mag2 += -1.0587
mb10364_mod_mag2 += -1.8983
ax1.errorbar(mb09260_data['t_phot1'], mb09260_mag1_rs,
ls='none', yerr=mb09260_data['mag_err1'],
color='tab:blue', alpha = 0.3)
ax1.errorbar(mb09260_data['t_phot2'], mb09260_mag2_rs,
ls='none', yerr=mb09260_data['mag_err2'],
color='tab:orange', alpha = 0.9)
ax1.plot(times_mod, mb09260_mod_mag1, '-', color = 'tab:blue')
ax1.plot(mb09260_data['t_phot1'], mb09260_mag1_rs, '.',
color='tab:blue', alpha = 0.3)
ax1.plot(1, 1, '.', color='tab:blue', alpha = 0.9, label='MOA I')
ax1.plot(mb09260_data['t_phot2'], mb09260_mag2_rs, 'o',
color='tab:orange', alpha = 0.9,
label='HST F814W - 1.06')
ax1.plot(times_mod, mb09260_mod_mag2, '--', color = 'tab:orange')
ax1.set_ylabel('Magnitude')
ax1.set_xlabel('Time (MJD)')
ax1.set_title('MB09260')
ax1.invert_yaxis()
ax1.set_xlim(54447, 56647)
ax1.xaxis.set_major_locator(plt.MultipleLocator(500))
ax1.yaxis.set_major_locator(plt.MultipleLocator(1))
ax1.set_ylim(18, 14)
ax1.legend()
ax2.errorbar(mb10364_data['t_phot1'], mb10364_mag1_rs,
ls='none', yerr=mb10364_data['mag_err1'],
color='tab:blue', alpha = 0.3)
ax2.errorbar(mb10364_data['t_phot2'], mb10364_mag2_rs,
ls='none', yerr=mb10364_data['mag_err2'],
color='tab:orange', alpha = 0.9)
ax2.plot(times_mod, mb10364_mod_mag1, '-', color = 'tab:blue')
ax2.plot(mb10364_data['t_phot1'], mb10364_mag1_rs, '.',
color='tab:blue', alpha = 0.3)
ax2.plot(1, 1, '.', color='tab:blue', alpha = 0.9, label='MOA I')
ax2.plot(mb10364_data['t_phot2'], mb10364_mag2_rs, 'o',
color='tab:orange', alpha = 0.9,
label='HST F606W - 1.90')
ax2.plot(times_mod, mb10364_mod_mag2, '--', color = 'tab:orange')
ax2.set_title('MB10364')
ax2.set_xlabel('Time (MJD)')
ax2.invert_yaxis()
ax2.set_xlim(54817, 56653)
ax2.xaxis.set_major_locator(plt.MultipleLocator(500))
ax2.yaxis.set_major_locator(plt.MultipleLocator(1))
ax2.set_ylim(15, 11.4)
ax2.legend()
plt.savefig('moa_hst_photometry.png')
def piE_tE_phot_only_fits():
span=0.999999426697
smooth=0.02
quantiles_2d=None
hist2d_kwargs=None
labels=None
label_kwargs=None
show_titles=False
title_fmt=".2f"
title_kwargs=None
"""
!!! NOTE: CHOICE OF THE quantiles_2d HAS A LARGE EFFECT
ON THE WAY THIS PLOT LOOKS !!!
Plot piE-tE 2D posteriors from OGLE photometry only fits.
Also plot PopSyCLE simulations simultaneously.
"""
# Initialize values.
if label_kwargs is None:
label_kwargs = dict()
if title_kwargs is None:
title_kwargs = dict()
if hist2d_kwargs is None:
hist2d_kwargs = dict()
# Set defaults.
hist2d_kwargs['alpha'] = hist2d_kwargs.get('alpha', 0.2)
hist2d_kwargs['levels'] = hist2d_kwargs.get('levels', quantiles_2d)
# MB09260 fit results.
data_09260 = munge.getdata2('mb09260',
phot_data=['MOA'],
ast_data = [])
fitter_09260 = model_fitter.PSPL_Solver(data_09260,
model.PSPL_Phot_Par_Param1,
outputfiles_basename = mb09260_dir)
results_09260 = fitter_09260.load_mnest_results_for_dynesty()
smy_09260 = fitter_09260.load_mnest_summary()
# MB10364 fit results.
data_10364 = munge.getdata2('mb10364',
phot_data=['MOA'],
ast_data = [])
fitter_10364 = model_fitter.PSPL_Solver(data_10364,
model.PSPL_Phot_Par_Param1,
outputfiles_basename = mb10364_dir)
results_10364 = fitter_10364.load_mnest_results_for_dynesty()
smy_10364 = fitter_10364.load_mnest_summary()
# OB110462 fit results.
data_110462 = munge.getdata2('ob110462',
phot_data=['I_OGLE'],
ast_data=[])
fitter_110462 = model_fitter.PSPL_Solver(data_110462,
model.PSPL_Phot_Par_Param1,
outputfiles_basename = ob110462_dir)
results_110462 = fitter_110462.load_mnest_results_for_dynesty()
smy_110462 = fitter_110462.load_mnest_summary()
# OB110310 fit results.
data_110310 = munge.getdata2('ob110310',
phot_data=['I_OGLE'],
ast_data=[])
fitter_110310 = model_fitter.PSPL_Solver(data_110310,
model.PSPL_Phot_Par_Param1,
outputfiles_basename = ob110310_dir)
results_110310 = fitter_110310.load_mnest_results_for_dynesty()
smy_110310 = fitter_110310.load_mnest_summary()
# OB110037 fit results.
data_110037 = munge.getdata2('ob110037',
phot_data=['I_OGLE'],
ast_data=[])
fitter_110037 = model_fitter.PSPL_Solver(data_110037,
model.PSPL_Phot_Par_Param1,
outputfiles_basename = ob110037_dir)
results_110037 = fitter_110037.load_mnest_results_for_dynesty()
smy_110037 = fitter_110037.load_mnest_summary()
# OB120169 fit results.
data_120169 = munge.getdata2('ob120169',
phot_data=['I_OGLE'],
ast_data=['Kp_Keck'])
fitter_120169 = model_fitter.PSPL_Solver(data_120169,
model.PSPL_Phot_Par_Param1,
add_error_on_photometry=True,
outputfiles_basename=ob120169_dir + 'c3_')
results_120169 = fitter_120169.load_mnest_results_for_dynesty()
smy_120169 = fitter_120169.load_mnest_summary()
# OB140613 fit results.
data_140613 = munge.getdata2('ob140613',
phot_data=['I_OGLE'],
ast_data=['Kp_Keck'])
fitter_140613 = model_fitter.PSPL_Solver(data_140613,
model.PSPL_Phot_Par_Param1,
multiply_error_on_photometry=True,
outputfiles_basename=ob140613_dir + 'c8_')
results_140613 = fitter_140613.load_mnest_results_for_dynesty()
smy_140613 = fitter_140613.load_mnest_summary()
# OB150029 fit results.
data_150029 = munge.getdata2('ob150029',
phot_data=['I_OGLE'],
ast_data=['Kp_Keck'])
fitter_150029 = model_fitter.PSPL_Solver(data_150029,
model.PSPL_Phot_Par_Param1,
add_error_on_photometry=True,
outputfiles_basename=ob150029_dir + 'd8_')
results_150029 = fitter_150029.load_mnest_results_for_dynesty()
smy_150029 = fitter_150029.load_mnest_summary()
# OB150211 fit results.
data_150211 = munge.getdata2('ob150211',
phot_data=['I_OGLE'],
ast_data=['Kp_Keck'])
fitter_150211 = model_fitter.PSPL_Solver(data_150211,
model.PSPL_Phot_Par_Param1,
add_error_on_photometry=True,
outputfiles_basename=ob150211_dir + 'a4_')
results_150211 = fitter_150211.load_mnest_results_for_dynesty()
smy_150211 = fitter_150211.load_mnest_summary()
# Extract weighted samples.
samples_09260 = results_09260['samples']
samples_10364 = results_10364['samples']
samples_110462 = results_110462['samples']
samples_110310 = results_110310['samples']
samples_110037 = results_110037['samples']
samples_120169 = results_120169['samples']
samples_140613 = results_140613['samples']
samples_150029 = results_150029['samples']
samples_150211 = results_150211['samples']
try:
weights_09260 = np.exp(results_09260['logwt'] - results_09260['logz'][-1])
weights_10364 = np.exp(results_10364['logwt'] - results_10364['logz'][-1])
weights_110462 = np.exp(results_110462['logwt'] - results_110462['logz'][-1])
weights_110310 = np.exp(results_110310['logwt'] - results_110310['logz'][-1])
weights_110037 = np.exp(results_110037['logwt'] - results_110037['logz'][-1])
weights_120169 = np.exp(results_120169['logwt'] - results_120169['logz'][-1])
weights_140613 = np.exp(results_140613['logwt'] - results_140613['logz'][-1])
weights_150029 = np.exp(results_150029['logwt'] - results_150029['logz'][-1])
weights_150211 = np.exp(results_150211['logwt'] - results_150211['logz'][-1])
except:
weights_09260 = results_09260['weights']
weights_10364 = results_10364['weights']
weights_110462 = results_110462['weights']
weights_110310 = results_110310['weights']
weights_110037 = results_110037['weights']
weights_120169 = results_120169['weights']
weights_140613 = results_140613['weights']
weights_150029 = results_150029['weights']
weights_150211 = results_150211['weights']
# Deal with 1D results. A number of extra catches are also here
# in case users are trying to plot other results besides the `Results`
# instance generated by `dynesty`.
samples_09260 = np.atleast_1d(samples_09260)
if len(samples_09260.shape) == 1:
samples_09260 = np.atleast_2d(samples_09260)
else:
assert len(samples_09260.shape) == 2, "Samples must be 1- or 2-D."
samples_09260 = samples_09260.T
assert samples_09260.shape[0] <= samples_09260.shape[1], "There are more " \
"dimensions than samples!"
samples_10364 = np.atleast_1d(samples_10364)
if len(samples_10364.shape) == 1:
samples_10364 = np.atleast_2d(samples_10364)
else:
assert len(samples_10364.shape) == 2, "Samples must be 1- or 2-D."
samples_10364 = samples_10364.T
assert samples_10364.shape[0] <= samples_10364.shape[1], "There are more " \
"dimensions than samples!"
samples_110462 = np.atleast_1d(samples_110462)
if len(samples_110462.shape) == 1:
samples_110462 = np.atleast_2d(samples_110462)
else:
assert len(samples_110462.shape) == 2, "Samples must be 1- or 2-D."
samples_110462 = samples_110462.T
assert samples_110462.shape[0] <= samples_110462.shape[1], "There are more " \
"dimensions than samples!"
samples_110310 = np.atleast_1d(samples_110310)
if len(samples_110310.shape) == 1:
samples_110310 = np.atleast_2d(samples_110310)
else:
assert len(samples_110310.shape) == 2, "Samples must be 1- or 2-D."
samples_110310 = samples_110310.T
assert samples_110310.shape[0] <= samples_110310.shape[1], "There are more " \
"dimensions than samples!"
samples_110037 = np.atleast_1d(samples_110037)
if len(samples_110037.shape) == 1:
samples_110037 = np.atleast_2d(samples_110037)
else:
assert len(samples_110037.shape) == 2, "Samples must be 1- or 2-D."
samples_110037 = samples_110037.T
assert samples_110037.shape[0] <= samples_110037.shape[1], "There are more " \
"dimensions than samples!"
samples_120169 = np.atleast_1d(samples_120169)
if len(samples_120169.shape) == 1:
samples_120169 = np.atleast_2d(samples_120169)
else:
assert len(samples_120169.shape) == 2, "Samples must be 1- or 2-D."
samples_120169 = samples_120169.T
assert samples_120169.shape[0] <= samples_120169.shape[1], "There are more " \
"dimensions than samples!"
samples_140613 = np.atleast_1d(samples_140613)
if len(samples_140613.shape) == 1:
samples_140613 = np.atleast_2d(samples_140613)
else:
assert len(samples_140613.shape) == 2, "Samples must be 1- or 2-D."
samples_140613 = samples_140613.T
assert samples_140613.shape[0] <= samples_140613.shape[1], "There are more " \
"dimensions than samples!"
samples_150029 = np.atleast_1d(samples_150029)
if len(samples_150029.shape) == 1:
samples_150029 = np.atleast_2d(samples_150029)
else:
assert len(samples_150029.shape) == 2, "Samples must be 1- or 2-D."
samples_150029 = samples_150029.T
assert samples_150029.shape[0] <= samples_150029.shape[1], "There are more " \
"dimensions than samples!"
samples_150211 = np.atleast_1d(samples_150211)
if len(samples_150211.shape) == 1:
samples_150211 = np.atleast_2d(samples_150211)
else:
assert len(samples_150211.shape) == 2, "Samples must be 1- or 2-D."
samples_150211 = samples_150211.T
assert samples_150211.shape[0] <= samples_150211.shape[1], "There are more " \
"dimensions than samples!"
# Plot the piE-tE 2D posteriors.
# tE = 2; piEE,N = 3, 4
fig, ax = plt.subplots(1, 2, figsize=(14, 6), sharey=True,
gridspec_kw={'width_ratios': [1.2, 2]})
plt.subplots_adjust(left=0.1, bottom=0.15, wspace=0.1, right=1)
tE_09260 = samples_09260[2]
tE_10364 = samples_10364[2]
tE_110462 = samples_110462[2]
tE_110310 = samples_110310[2]
tE_110037 = samples_110037[2]
tE_120169 = samples_120169[2]
tE_140613 = samples_140613[2]
tE_150029 = samples_150029[2]
tE_150211 = samples_150211[2]
piE_09260 = np.hypot(samples_09260[3], samples_09260[4])
piE_10364 = np.hypot(samples_10364[3], samples_10364[4])
piE_110462 = np.hypot(samples_110462[3], samples_110462[4])
piE_110310 = np.hypot(samples_110310[3], samples_110310[4])
piE_110037 = np.hypot(samples_110037[3], samples_110037[4])
piE_120169 = np.hypot(samples_120169[3], samples_120169[4])
piE_140613 = np.hypot(samples_140613[3], samples_140613[4])
piE_150029 = np.hypot(samples_150029[3], samples_150029[4])
piE_150211 = np.hypot(samples_150211[3], samples_150211[4])
sx = smooth
sy = smooth
hist2d_kwargs['fill_contours'] = hist2d_kwargs.get('fill_contours',
False)
hist2d_kwargs['plot_contours'] = hist2d_kwargs.get('plot_contours',
True)
model_fitter.contour2d_alpha(tE_09260, piE_09260, span=[span, span], quantiles_2d=quantiles_2d,
weights=weights_09260, ax=ax[1], smooth=[sy, sx], color='blue',
**hist2d_kwargs, plot_density=False)
model_fitter.contour2d_alpha(tE_110310, piE_110310, span=[span, span], quantiles_2d=quantiles_2d,
weights=weights_110310, ax=ax[1], smooth=[sy, sx], color='blue',
**hist2d_kwargs, plot_density=False)
model_fitter.contour2d_alpha(tE_120169, piE_120169, span=[span, span], quantiles_2d=quantiles_2d,
weights=weights_120169, ax=ax[1], smooth=[sy, sx], color='red',
**hist2d_kwargs, plot_density=False)
model_fitter.contour2d_alpha(tE_150211, piE_150211, span=[span, span], quantiles_2d=quantiles_2d,
weights=weights_150211, ax=ax[1], smooth=[sy, sx], color='red',
**hist2d_kwargs, plot_density=False)
# Maximum likelihood vals
smy_list = [smy_09260, smy_10364, smy_110037,
smy_110310, smy_110462, smy_120169,
smy_140613, smy_150029, smy_150211]
smy_name = ['MB09260', 'MB10364', 'OB110037',
'OB110310', 'OB110462', 'OB120169',
'OB140613', 'OB150029', 'OB150211']
maxl = {}
for ss, smy in enumerate(smy_list):
print(smy_name[ss])
print('t0 : ', smy['MaxLike_t0'][0])
print('tE : ', smy['MaxLike_tE'][0])
print('piE : ', np.hypot(smy['MaxLike_piE_E'][0], smy['MaxLike_piE_N'][0]))
maxl[smy_name[ss]] = {'tE' : smy['MaxLike_tE'][0],
'piE' : np.hypot(smy['MaxLike_piE_E'][0], smy['MaxLike_piE_N'][0])}
plt.plot(maxl['MB10364']['tE'], maxl['MB10364']['piE'], color='blue', marker='s')
plt.plot(maxl['OB110037']['tE'], maxl['OB110037']['piE'], color='blue', marker='s')
plt.plot(maxl['OB110462']['tE'], maxl['OB110462']['piE'], color='blue', marker='s', label = 'HST')
plt.plot(61.4, np.hypot(-0.393, -0.071), color = 'red', marker='^') # OB110022
plt.plot(maxl['OB140613']['tE'], maxl['OB140613']['piE'], color='red', marker='^')
plt.plot(maxl['OB150029']['tE'], maxl['OB150029']['piE'], color='red', marker='^', label = 'Keck')
# Add the PopSyCLE simulation points.
# NEED TO UPDATE THIS WITH BUGFIX IN DELTAM
t = Table.read('/u/casey/scratch/papers/microlens_2019/popsycle_rr_files/Mock_EWS_v2.fits')
bh_idx = np.where(t['rem_id_L'] == 103)[0]
ns_idx = np.where(t['rem_id_L'] == 102)[0]
wd_idx = np.where(t['rem_id_L'] == 101)[0]
st_idx = np.where(t['rem_id_L'] == 0)[0]
u0_arr = t['u0']
thetaE_arr = t['theta_E']
# Stores the maximum astrometric shift
final_delta_arr = np.zeros(len(u0_arr))
# Stores the lens-source separation corresponding
# to the maximum astrometric shift
final_u_arr = np.zeros(len(u0_arr))
# Sort by whether the maximum astrometric shift happens
# before or after the maximum photometric amplification
big_idx = np.where(u0_arr > np.sqrt(2))[0]
small_idx = np.where(u0_arr <= np.sqrt(2))[0]
# Flux ratio of lens to source (and make it 0 if dark lens)
g_arr = 10**(-0.4 * (t['ubv_i_app_L'] - t['ubv_i_app_S']))
g_arr = np.nan_to_num(g_arr)
for i in np.arange(len(u0_arr)):
g = g_arr[i]
thetaE = thetaE_arr[i]
# Try all values between u0 and sqrt(2) to find max
# astrometric shift
if u0_arr[i] < np.sqrt(2):
u_arr = np.linspace(u0_arr[i], np.sqrt(2), 100)
delta_arr = np.zeros(len(u_arr))
for j in np.arange(len(u_arr)):
u = u_arr[j]
numer = 1 + g * (u**2 - u * np.sqrt(u**2 + 4) + 3)
denom = u**2 + 2 + g * u * np.sqrt(u**2 + 4)
delta = (u * thetaE/(1 + g)) * (numer/denom)
delta_arr[j] = delta
max_idx = np.argmax(delta_arr)
final_delta_arr[i] = delta_arr[max_idx]
final_u_arr[i] = u_arr[max_idx]
# Maximum astrometric shift will occur at sqrt(2)
if u0_arr[i] > np.sqrt(2):
u = u0_arr[i]
numer = 1 + g * (u**2 - u * np.sqrt(u**2 + 4) + 3)
denom = u**2 + 2 + g * u * np.sqrt(u**2 + 4)
delta = (u * thetaE/(1 + g)) * (numer/denom)
final_delta_arr[i] = delta
final_u_arr[i] = u
ax[0].scatter(final_delta_arr[st_idx], t['pi_E'][st_idx],
alpha = 0.4, marker = '.', s = 25,
c = 'gold')
ax[0].scatter(final_delta_arr[wd_idx], t['pi_E'][wd_idx],
alpha = 0.4, marker = '.', s = 25,
c = 'goldenrod')
ax[0].scatter(final_delta_arr[ns_idx], t['pi_E'][ns_idx],
alpha = 0.4, marker = '.', s = 25,
c = 'sienna')
ax[0].scatter(final_delta_arr[bh_idx], t['pi_E'][bh_idx],
alpha = 0.8, marker = '.', s = 25,
c = 'black')
ax[1].scatter(t['t_E'][st_idx], t['pi_E'][st_idx],
alpha = 0.4, marker = '.', s = 25,
color = 'gold')
ax[1].scatter(t['t_E'][wd_idx], t['pi_E'][wd_idx],
alpha = 0.4, marker = '.', s = 25,
color = 'goldenrod')
ax[1].scatter(t['t_E'][ns_idx], t['pi_E'][ns_idx],
alpha = 0.4, marker = '.', s = 25,
color = 'sienna')
ax[1].scatter(t['t_E'][bh_idx], t['pi_E'][bh_idx],
alpha = 0.8, marker = '.', s = 25,
color = 'black')
# Trickery to make the legend darker
ax[1].scatter(0.01, 100,
alpha = 0.8, marker = 'o', s = 25,
label = 'Star', color = 'gold')
ax[1].scatter(0.01, 100,
alpha = 0.8, marker = 'o', s = 25,
label = 'WD', color = 'goldenrod')
ax[1].scatter(0.01, 100,
alpha = 0.8, marker = 'o', s = 25,
label = 'NS', color = 'sienna')
ax[1].scatter(0.01, 100,
alpha = 0.8, marker = 'o', s = 25,
label = 'BH', color = 'black')
ax[0].set_xlabel('$\delta_{c,max}$ (mas)')
ax[0].set_ylabel('$\pi_E$')
ax[0].set_xscale('log')
ax[0].set_yscale('log')
ax[1].set_xlim(10, 400)
ax[1].set_ylim(0.009, 0.5)
ax[1].set_xlabel('$t_E$ (days)')
ax[1].set_xscale('log')
ax[1].set_yscale('log')
box = ax[1].get_position()
ax[1].set_position([box.x0, box.y0, box.width * 0.7, box.height])
ax[1].legend(bbox_to_anchor=(1.4, 0.5), loc="center right")
plt.savefig('hst_sahu_piE_tE_phot_only_fit.png')
def modelfit(target,
align_dir,
phot_only=False,
parallax=False,
solve=True,
points_dir='points_d/',
runcode='aa_'):
data = getdata(target, align_dir + points_dir)
if parallax == False:
if phot_only:
mdir = 'mnest_pspl_phot/'
fit = model_fitter.PSPL_phot_Solver(data)
if not phot_only:
mdir = 'mnest_pspl/'
fit = model_fitter.PSPL_Solver(data)
elif parallax == True:
if phot_only:
mdir = 'mnest_pspl_par_phot/'
fit = model_fitter.PSPL_phot_parallax_Solver(data)
if not phot_only:
mdir = 'mnest_pspl_par/'
fit = model_fitter.PSPL_parallax_Solver(data)
fit.outputfiles_basename = align_dir + mdir + runcode
if solve:
if target == 'ob150211':
fit.mag_base_gen = model_fitter.make_gen(16.0, 18.0)
fit.dL_gen = model_fitter.make_gen(500, 8000)
if not os.path.exists(align_dir + mdir):
os.mkdir(align_dir + mdir)
fit.solve()
fit.plot_posteriors()
if not os.path.exists(align_dir + mdir):
raise ValueError('This model has not been ran yet.')
modeled = fit.get_best_fit_model()
# Create t_dat vector of times covering data times + interim periods to get a smooth photometric model
t_dat = np.linspace(np.min(data['t_phot']),
np.max(data['t_phot']),
num=len(data['t_phot']) * 2,
endpoint=True)
mag_dat = modeled.get_photometry(t_dat)
mag = modeled.get_photometry(data['t_phot'])
lnL_phot = modeled.likely_photometry(data['t_phot'], data['mag'],
data['mag_err'])
if phot_only:
lnL = lnL_phot.mean()
if not phot_only:
pos = modeled.get_astrometry(data['t_ast'])
lnL_ast = modeled.likely_astrometry(data['t_ast'], data['xpos'],
data['ypos'], data['xpos_err'],
data['ypos_err'])
lnL = lnL_phot.mean() + lnL_ast.mean()
print('lnL: ', lnL)
if not os.path.exists(align_dir + mdir + 'plots/'):
os.mkdir(align_dir + mdir + 'plots/')
fig1, (pho, pho_res) = py.subplots(2,
1,
figsize=(10, 10),
gridspec_kw={'height_ratios': [3, 1]},
sharex=True)
fig1.subplots_adjust(hspace=0)
pho.errorbar(data['t_phot'],
data['mag'],
yerr=data['mag_err'],
fmt='k.',
label='OGLE-IV')
pho.plot(t_dat, mag_dat, 'r-', label='best-fit model')
pho.set_ylabel('mag')
pho.invert_yaxis()
pho.legend()
pho.set_title(target + ' Photometry')
pho_res.errorbar(data['t_phot'],
data['mag'] - mag,
yerr=data['mag_err'],
fmt='k.')
pho_res.plot(data['t_phot'], mag - mag, 'r-', lw=2)
pho_res.invert_yaxis()
pho_res.set_ylabel('data - model')
pho_res.set_xlabel('days (MJD)')
py.savefig(align_dir + mdir + 'plots/photo.png', bbox_inches='tight')
if not phot_only:
fig2, (x, x_res) = py.subplots(2,
1,
gridspec_kw={'height_ratios': [3, 1]},
sharex=True)
fig2.subplots_adjust(hspace=0)
x.errorbar(data['t_ast'],
data['xpos'],
yerr=data['xpos_err'],
fmt='k.',
label='aligned data')
x.plot(data['t_ast'], pos[:, 0], 'r-', label='model')
x.set_ylabel('X Pos (")')
x.legend()
x_res.errorbar(data['t_ast'],
data['xpos'] - pos[:, 0],
yerr=data['xpos_err'],
fmt='k.')
x_res.plot(data['t_ast'], pos[:, 0] - pos[:, 0], 'r-')
x_res.set_xlabel('days (MJD)')
x_res.set_ylabel('data - model')
x.set_title('X')
py.savefig(align_dir + mdir + 'plots/x_pos.png', bbox_inches='tight')
fig3, (y, y_res) = py.subplots(2,
1,
gridspec_kw={'height_ratios': [3, 1]},
sharex=True)
fig3.subplots_adjust(hspace=0)
y.errorbar(data['t_ast'],
data['ypos'],
yerr=data['ypos_err'],
fmt='k.',
label='aligned data')
y.plot(data['t_ast'], pos[:, 1], 'r-', label='model')
y.set_ylabel('Y Pos (")')
y.legend()
y_res.errorbar(data['t_ast'],
data['ypos'] - pos[:, 1],
yerr=data['ypos_err'],
fmt='k.')
y_res.plot(data['t_ast'], pos[:, 1] - pos[:, 1], 'r-')
y_res.set_xlabel('days (MJD)')
y_res.set_ylabel('data - model')
y.set_title('Y')
py.savefig(align_dir + mdir + 'plots/y_pos.png', bbox_inches='tight')
py.figure(4)
py.clf()
py.errorbar(data['xpos'],
data['ypos'],
xerr=data['xpos_err'],
yerr=data['ypos_err'],
fmt='k.')
py.plot(pos[:, 0], pos[:, 1], 'r-')
py.gca().invert_xaxis()
py.xlabel('X Pos (")')
py.ylabel('Y Pos (")')
py.legend(['model', 'aligned data'])
py.title(target + ' X and Y')
py.savefig(align_dir + mdir + 'plots/pos.png', bbox_inches='tight')
fit.summarize_results()
best = fit.get_best_fit()
out = open(align_dir + mdir + runcode + 'final.txt', 'w')
pars, q = quantiles(fit)
out.write(' best median\n')
for n in pars:
out.write('%15s %10.3f %10.3f + %10.3f - %10.3f\n' % \
(n, best[n], q[n][0], q[n][1], q[n][2]))
out.close()
return