def make_figure(KOI,
planet_number,
results,
t,
y,
y_filt,
trend,
rawtime,
rawflux,
catalog,
row,
valid=False):
fig = plt.figure(figsize=(10, 14))
G = gridspec.GridSpec(9, 3)
G.update(hspace=0)
radius = catalog["R_star"][row]
radius_max = catalog["rad_up"][row]
radius_min = catalog["rad_down"][row]
mass = catalog["mass"][row]
mass_max = catalog["mass_up"][row]
mass_min = catalog["mass_down"][row]
# Raw flux
axes_1a = plt.subplot(G[0:1, :2])
plt.plot(rawtime, rawflux / numpy.mean(rawflux), "k", linewidth=0.5)
plt.plot(rawtime, trend / numpy.mean(y), color='red', linewidth=0.5)
plt.xlim(min(t), max(t))
plt.xticks(())
plt.ylabel(r'Raw Flux')
ppm_flux = -(1 - y_filt) * 10**6
# Detrended flux
G = gridspec.GridSpec(9, 3)
G.update(hspace=0.4)
axes_1b = plt.subplot(G[1:2, :2])
plt.plot(t, ppm_flux, "k", linewidth=0.5)
top_ppm = -(1 - max(y_filt)) * 10**6
bottom_ppm = -(1 - min(y_filt)) * 10**6
y_range = abs(bottom_ppm) + abs(top_ppm)
x_range = max(t) - min(t)
offset_y = bottom_ppm + y_range / 40
offset_x = min(t) + x_range / 80
std = numpy.std(ppm_flux)
text = r'$\sigma=$' + format(std, '.0f') + r'$\,$ppm'
plt.text(offset_x, offset_y, text, color='red')
plt.ylabel(r'Flux (ppm)')
plt.xlabel('Time (BKJD, days)')
plt.xlim(min(t), max(t))
# Phase fold
axes_2 = plt.subplot(G[2, 0:2])
plt.plot((results.model_folded_phase - 0.5) * results.period,
-(1 - results.model_folded_model) * 10**6,
color='red',
zorder=99)
plt.scatter((results.folded_phase - 0.5) * results.period,
-(1 - results.folded_y) * 10**6,
color='black',
s=2,
alpha=0.5,
zorder=2)
plt.xlim(-2 * results.duration, 2 * results.duration)
plt.xlabel('Time from mid-transit (days)')
plt.ylabel('Flux')
plt.ylim(
(numpy.percentile(-(1 - y_filt) * 10**6,
0.1), numpy.percentile(-(1 - y_filt) * 10**6, 99.5)))
plt.text(-1.95 * results.duration,
numpy.percentile(-(1 - y_filt) * 10**6, 0.1), 'primary')
plt.plot((results.folded_phase - 0.5) * results.period,
-(1 - running_mean_equal_length(results.folded_y, 10)) * 10**6,
color='blue',
linestyle='dashed',
linewidth=1,
zorder=3)
# Phase fold to secondary eclipse
G = gridspec.GridSpec(9, 3)
G.update(hspace=0.3)
axes_3 = plt.subplot(G[2, 2])
plt.yticks(())
phases = fold(time=t, period=results.period, T0=results.T0)
sort_index = numpy.argsort(phases, kind="mergesort")
phases = phases[sort_index]
flux = y_filt[sort_index]
plt.scatter((phases - 0.5) * results.period,
flux,
color='black',
s=2,
alpha=0.5,
zorder=2)
plt.plot((-0.5, 0.5), (1, 1), color='red')
plt.xlim(-2 * results.duration, 2 * results.duration)
plt.ylim(numpy.percentile(y_filt, 0.1), numpy.percentile(y_filt, 99.5))
plt.plot((phases - 0.5) * results.period,
running_mean_equal_length(flux, 10),
color='blue',
linestyle='dashed',
linewidth=1,
zorder=3)
# Calculate secondary eclipse depth
intransit_secondary = numpy.where(
numpy.logical_and(
(phases - 0.5) * results.period > (-0.5 * results.duration),
(phases - 0.5) * results.period < (0.5 * results.duration)))
mean = -(1 - numpy.mean(flux[intransit_secondary])) * 10**6
stabw = -(numpy.std(flux[intransit_secondary]) /
numpy.sqrt(len(flux[intransit_secondary]))) * 10**6
significance_eb = mean / stabw
plt.scatter((phases[intransit_secondary] - 0.5) * results.period,
flux[intransit_secondary],
color='orange',
s=20,
alpha=0.5,
zorder=0)
if numpy.isnan(mean):
mean = 0
if numpy.isnan(stabw):
stabw = 0
if numpy.isnan(significance_eb):
significance_eb = 99
text = r'secondary ' + str(
int(mean)) + r'$\pm$' + str(int(stabw)) + ' ppm (' + format(
significance_eb, '.1f') + r'$\,\sigma$)'
plt.text(-1.95 * results.duration, numpy.percentile(y_filt, 0.1), text)
# Full phase
G = gridspec.GridSpec(9, 3)
G.update(hspace=1)
axes_5 = plt.subplot(G[3, :])
plt.plot(results.model_folded_phase,
-(1 - results.model_folded_model) * 10**6,
color='red')
plt.scatter(results.folded_phase,
-(1 - results.folded_y) * 10**6,
color='black',
s=2,
alpha=0.5,
zorder=2)
plt.xlim(0, 1)
plt.ylim(numpy.percentile(ppm_flux, 0.1), numpy.percentile(ppm_flux, 99.9))
plt.xlabel('Phase')
plt.ylabel('Flux')
# Check if phase gaps
phase_diffs = abs(results.folded_phase -
numpy.roll(results.folded_phase, -1))
phase_diffs = phase_diffs[:-1]
largest_phase_peak = max(phase_diffs) / numpy.median(phase_diffs)
# All transits in time series
G = gridspec.GridSpec(9, 3)
G.update(hspace=0)
axes_6b = plt.subplot(G[4, :])
y_filt = -(1 - y_filt) * 10**6
in_transit = transit_mask(t, results.period, results.duration, results.T0)
t_diff1 = abs(t - numpy.roll(t, -1))
t_diff2 = abs(t - numpy.roll(t, +1))
if max(t_diff1[in_transit]) > 0.1 or max(
t_diff2[in_transit]) > 0.1: # in days
transit_near_gaps = True
plt.text(min(t), numpy.percentile(ppm_flux, 0.1),
'Warning: Transits near gaps')
else:
transit_near_gaps = False
transit_touches_edge = False
if max(t) in t[in_transit]:
plt.text(min(t), numpy.percentile(ppm_flux, 0.1),
'Warning: Last transit touches end')
transit_touches_edge = True
if max(t) in t[in_transit]:
plt.text(min(t), numpy.percentile(ppm_flux, 0.1),
'Warning: First transit touches start')
transit_touches_edge = True
plt.scatter(t[in_transit], y_filt[in_transit], color='red', s=2, zorder=0)
plt.scatter(t[~in_transit],
y_filt[~in_transit],
color='black',
alpha=0.5,
s=2,
zorder=0)
plt.plot(results.model_lightcurve_time,
-(1 - results.model_lightcurve_model) * 10**6,
alpha=0.5,
color='red',
zorder=1)
plt.xlim(min(t), max(t))
plt.ylim(numpy.percentile(ppm_flux, 0.1), numpy.percentile(ppm_flux, 99.9))
plt.xlabel('Flux')
plt.ylabel('Flux')
plt.xticks(())
# Transit depths error bars
avg = -(1 - results.depth_mean[0]) * 10**6
if numpy.isnan(results.transit_depths_uncertainties).any():
step = 0
else:
step = max(results.transit_depths_uncertainties) * 10**6
down = avg - step
G = gridspec.GridSpec(9, 3)
G.update(hspace=1)
axes_6 = plt.subplot(G[5, :])
plt.errorbar(results.transit_times,
-(1 - results.transit_depths) * 10**6,
yerr=step,
fmt='o',
color='red')
plt.plot((min(t), max(t)), (0, 0), color='black')
plt.plot((min(t), max(t)), (avg, avg), color='black', linestyle='dashed')
plt.xlim(min(t), max(t))
for transit in range(len(results.transit_times)):
plt.text(results.transit_times[transit],
down,
str(int(results.per_transit_count[transit])),
horizontalalignment='center')
plt.xlabel('Time (BKJD, days)')
plt.ylabel('Flux')
# Test statistic
G = gridspec.GridSpec(9, 3)
G.update(hspace=0)
#G.update(hspace=3)
axes_7 = plt.subplot(G[6, :])
axes_7.axvline(results.period, alpha=0.4, lw=3)
for n in range(2, 5):
axes_7.axvline(n * results.period, alpha=0.4, lw=1, linestyle="dashed")
axes_7.axvline(results.period / n, alpha=0.4, lw=1, linestyle="dashed")
plt.ylabel(r'TLS SDE')
plt.xticks(())
plt.plot(results.periods, results.power, color='black', lw=0.5)
plt.xlim(min(results.periods), max(results.periods))
plt.text(results.period + 0.1, results.SDE * 0.9,
'SDE=' + format(results.SDE, '.1f'))
# LS periodogram raw
freqs = numpy.geomspace(1 / min(results.periods), 1 / max(results.periods),
10000)
freqs = numpy.unique(freqs)
lspower = LombScargle(rawtime, rawflux).power(freqs)
freqs = 1 / freqs
G = gridspec.GridSpec(9, 3)
G.update(hspace=0)
axes_8 = plt.subplot(G[7, :])
axes_8.axvline(results.period, alpha=0.4, lw=15)
plt.ylabel(r'LS (raw)')
plt.plot(freqs, lspower, color='black', lw=0.5)
plt.xlim(min(results.periods), max(results.periods))
plt.ylim(0, max(lspower) * 1.1)
plt.xlabel('Period (days)')
# Calculate Lomb-Scargle periodogram detrended
freqs = numpy.geomspace(1 / min(results.periods), 1 / max(results.periods),
10000)
freqs = numpy.unique(freqs)
lspower = LombScargle(t, y_filt).power(freqs)
freqs = 1 / freqs
continuum = numpy.std(lspower)
lspower = numpy.divide(lspower, continuum) # Normalize to S/N
G = gridspec.GridSpec(9, 3)
G.update(hspace=0)
axes_9 = plt.subplot(G[8, :])
peak_index = numpy.argmax(lspower)
axes_9.axvline(freqs[peak_index], alpha=0.4, lw=10)
plt.ylabel(r'LS (detrended)')
plt.plot(freqs, lspower, color='black', lw=0.5)
plt.xlim(min(results.periods), max(results.periods))
plt.ylim(0, max(lspower) * 1.1)
plt.xlabel('Period (days)')
# Text
G = gridspec.GridSpec(9, 3)
G.update(hspace=0, wspace=0.1)
axes_8 = plt.subplot(G[0:2, 2])
plt.xticks(())
plt.yticks(())
plt.text(0,
0,
'KOI ' + str(KOI) + '.0' + str(planet_number),
fontweight='bold')
plt.text(
0, -0.30, 'SDE=' + format(results.SDE, '.1f') + ', SNR=' +
format(results.snr, '.1f'))
plt.text(
0, -0.45, 'P=' + format(results.period, '.5f') + ' +-' +
format(results.period_uncertainty, '.5f') + r'$\,$d')
plt.text(0, -0.60,
r'$T_{dur}=$' + format(results.duration, '.5f') + r'$\,$d')
plt.text(
0, -0.75,
r'$R_*=$' + format(radius, '.2f') + ' (+' + format(radius_max, '.2f') +
', -' + format(radius_min, '.2f') + ') $\,R_{\odot}$')
plt.text(
0, -0.90,
r'$M_*=$' + format(mass, '.2f') + ' (+' + format(mass_max, '.2f') +
', -' + format(mass_min, '.2f') + ') $\,M_{\odot}$')
plt.text(0, -1.05, r'$R_P/R_*=$' + format(results.rp_rs, '.3f'))
print('rp_rs', results.rp_rs)
rp = results.rp_rs * radius # in solar radii
sun = 695700
earth = 6371
jupi = 69911
rp_earth = (sun / earth) * rp
rp_jupi = (sun / jupi) * rp
plt.text(
0, -1.20, r'$R_P=$' + format(rp_earth, '.2f') + '$\,R_{\oplus}=$' +
format(rp_jupi, '.2f') + '$\,R_{Jup}$')
plt.text(
0, -1.35,
r'$\delta=$' + \
format((1-results.depth_mean[0])*10**6, '.0f') + ' ppm (' + \
format((1-results.depth_mean_odd[0])*10**6, '.0f') + ', ' + \
format((1-results.depth_mean_even[0])*10**6, '.0f') + ')')
plt.text(
0, -1.50, r'odd/even mismatch: ' +
format(results.odd_even_mismatch, '.2f') + '$\,\sigma$')
plt.text(
0, -1.65,
str(results.distinct_transit_count) + '/' +
str(results.transit_count) + ' transits with data')
plt.xlim(-0.1, 2)
plt.ylim(-2, 0.1)
axes_8.axis('off')
plt.subplots_adjust(hspace=0.4)
if abs(significance_eb) > 3:
valid = False
print('Vetting fail! significance_eb > 3')
if abs(results.odd_even_mismatch) > 3:
valid = False
print('Vetting fail! odd_even_mismatch larger 3 sigma')
if transit_near_gaps and results.distinct_transit_count < 4:
valid = False
print('Vetting fail! Transit near gaps and distinct_transit_count < 4')
if valid: # 5-50d
figure_out_path = 'GOOD0-5d_' + str(KOI) + '_0' + str(planet_number)
else:
figure_out_path = '_FAIL0-5d_' + str(KOI) + '_0' + str(planet_number)
plt.savefig(figure_out_path + '.png', bbox_inches='tight')
print('Figure made:', figure_out_path)
plt.close
return valid
def make_figure(KOI, planet_number, results, t, y, y_filt, trend, rawtime,
rawflux):
fig = plt.figure(figsize=(10, 14))
G = gridspec.GridSpec(9, 3)
G.update(hspace=0)
radius = catalog["R_star"][row]
radius_max = catalog["rad_up"][row]
radius_min = catalog["rad_down"][row]
mass = catalog["mass"][row]
mass_max = catalog["mass_up"][row]
mass_min = catalog["mass_down"][row]
# Raw flux
axes_1a = plt.subplot(G[0:1, :2])
plt.plot(rawtime, rawflux / numpy.mean(rawflux), "k", linewidth=0.5)
plt.plot(rawtime, trend / numpy.mean(y), color='red', linewidth=0.5)
plt.xlim(min(t), max(t))
plt.xticks(())
plt.ylabel(r'Raw Flux')
ppm_flux = -(1 - y_filt) * 10**6
# Detrended flux
G = gridspec.GridSpec(9, 3)
G.update(hspace=0.4)
axes_1b = plt.subplot(G[1:2, :2])
plt.plot(t, ppm_flux, "k", linewidth=0.5)
top_ppm = -(1 - max(y_filt)) * 10**6
bottom_ppm = -(1 - min(y_filt)) * 10**6
y_range = abs(bottom_ppm) + abs(top_ppm)
x_range = max(t) - min(t)
offset_y = bottom_ppm + y_range / 40
offset_x = min(t) + x_range / 80
#print(top_ppm, bottom_ppm, y_range, offset_y)
std = numpy.std(ppm_flux)
text = r'$\sigma=$' + format(std, '.0f') + r'$\,$ppm'
#print(bottom_ppm, top_ppm, offset, offset)
plt.text(offset_x, offset_y, text, color='red')
plt.ylabel(r'Flux (ppm)')
plt.xlabel('Time (BKJD, days)')
plt.xlim(min(t), max(t))
# Phase fold
axes_2 = plt.subplot(G[2, 0:2])
plt.plot((results.model_folded_phase - 0.5) * results.period,
-(1 - results.model_folded_model) * 10**6,
color='red',
zorder=99)
plt.scatter((results.folded_phase - 0.5) * results.period,
-(1 - results.folded_y) * 10**6,
color='black',
s=2,
alpha=0.5,
zorder=2)
plt.xlim(-2 * results.duration, 2 * results.duration)
plt.xlabel('Time from mid-transit (days)')
plt.ylabel('Flux')
plt.ylim(
(numpy.percentile(-(1 - y_filt) * 10**6,
0.1), numpy.percentile(-(1 - y_filt) * 10**6, 99.5)))
plt.text(-1.95 * results.duration,
numpy.percentile(-(1 - y_filt) * 10**6, 0.1), 'primary')
plt.plot((results.folded_phase - 0.5) * results.period,
-(1 - running_mean_equal_length(results.folded_y, 10)) * 10**6,
color='blue',
linestyle='dashed',
linewidth=1,
zorder=3)
# Phase fold to secondary eclipse
G = gridspec.GridSpec(9, 3)
G.update(hspace=0.3)
axes_3 = plt.subplot(G[2, 2])
plt.yticks(())
phases = fold(time=t, period=results.period, T0=results.T0)
sort_index = numpy.argsort(phases, kind="mergesort")
phases = phases[sort_index]
flux = y_filt[sort_index]
plt.scatter((phases - 0.5) * results.period,
flux,
color='black',
s=2,
alpha=0.5,
zorder=2)
plt.plot((-0.5, 0.5), (1, 1), color='red')
plt.xlim(-2 * results.duration, 2 * results.duration)
plt.ylim(numpy.percentile(y_filt, 0.1), numpy.percentile(y_filt, 99.5))
plt.plot((phases - 0.5) * results.period,
running_mean_equal_length(flux, 10),
color='blue',
linestyle='dashed',
linewidth=1,
zorder=3)
# Calculate secondary eclipse depth
intransit_secondary = numpy.where(
numpy.logical_and(
(phases - 0.5) * results.period > (-0.5 * results.duration),
(phases - 0.5) * results.period < (0.5 * results.duration)))
mean = -(1 - numpy.mean(flux[intransit_secondary])) * 10**6
stabw = -(numpy.std(flux[intransit_secondary]) /
numpy.sqrt(len(flux[intransit_secondary]))) * 10**6
significance_eb = mean / stabw
#print(mean, stabw, significance)
plt.scatter((phases[intransit_secondary] - 0.5) * results.period,
flux[intransit_secondary],
color='orange',
s=20,
alpha=0.5,
zorder=0)
if numpy.isnan(mean):
mean = 0
if numpy.isnan(stabw):
stabw = 0
if numpy.isnan(significance_eb):
significance_eb = 99
text = r'secondary ' + str(
int(mean)) + r'$\pm$' + str(int(stabw)) + ' ppm (' + format(
significance_eb, '.1f') + r'$\,\sigma$)'
plt.text(-1.95 * results.duration, numpy.percentile(y_filt, 0.1), text)
print('secondary vespa', (mean - stabw) * 10**-6)
# Full phase
G = gridspec.GridSpec(9, 3)
G.update(hspace=1)
axes_5 = plt.subplot(G[3, :])
plt.plot(results.model_folded_phase,
-(1 - results.model_folded_model) * 10**6,
color='red')
plt.scatter(results.folded_phase,
-(1 - results.folded_y) * 10**6,
color='black',
s=2,
alpha=0.5,
zorder=2)
plt.xlim(0, 1)
plt.ylim(numpy.percentile(ppm_flux, 0.1), numpy.percentile(ppm_flux, 99.9))
plt.xlabel('Phase')
plt.ylabel('Flux')
# Check if phase gaps
phase_diffs = abs(results.folded_phase -
numpy.roll(results.folded_phase, -1))
phase_diffs = phase_diffs[:-1]
largest_phase_peak = max(phase_diffs) / numpy.median(phase_diffs)
# All transits in time series
G = gridspec.GridSpec(9, 3)
G.update(hspace=0)
axes_6b = plt.subplot(G[4, :])
y_filt = -(1 - y_filt) * 10**6
in_transit = transit_mask(t, results.period, results.duration, results.T0)
t_diff1 = abs(t - numpy.roll(t, -1))
t_diff2 = abs(t - numpy.roll(t, +1))
if max(t_diff1[in_transit]) > 0.1 or max(
t_diff2[in_transit]) > 0.1: # in days
transit_near_gaps = True
plt.text(min(t), numpy.percentile(ppm_flux, 0.1),
'Warning: Transits near gaps')
else:
transit_near_gaps = False
transit_touches_edge = False
if max(t) in t[in_transit]:
plt.text(min(t), numpy.percentile(ppm_flux, 0.1),
'Warning: Last transit touches end')
transit_touches_edge = True
if max(t) in t[in_transit]:
plt.text(min(t), numpy.percentile(ppm_flux, 0.1),
'Warning: First transit touches start')
transit_touches_edge = True
plt.scatter(t[in_transit], y_filt[in_transit], color='red', s=2, zorder=0)
plt.scatter(t[~in_transit],
y_filt[~in_transit],
color='black',
alpha=0.5,
s=2,
zorder=0)
plt.plot(results.model_lightcurve_time,
-(1 - results.model_lightcurve_model) * 10**6,
alpha=0.5,
color='red',
zorder=1)
plt.xlim(min(t), max(t))
plt.ylim(numpy.percentile(ppm_flux, 0.1), numpy.percentile(ppm_flux, 99.9))
plt.xlabel('Flux')
plt.ylabel('Flux')
plt.xticks(())
# Transit depths error bars
avg = -(1 - results.depth_mean[0]) * 10**6
if numpy.isnan(results.transit_depths_uncertainties).any():
step = 0
else:
step = max(results.transit_depths_uncertainties) * 10**6
down = avg - step
G = gridspec.GridSpec(9, 3)
G.update(hspace=1)
axes_6 = plt.subplot(G[5, :])
plt.errorbar(results.transit_times,
-(1 - results.transit_depths) * 10**6,
yerr=step,
fmt='o',
color='red')
plt.plot((min(t), max(t)), (0, 0), color='black')
plt.plot((min(t), max(t)), (avg, avg), color='black', linestyle='dashed')
plt.xlim(min(t), max(t))
for transit in range(len(results.transit_times)):
plt.text(results.transit_times[transit],
down,
str(int(results.per_transit_count[transit])),
horizontalalignment='center')
#print(str(int(results.per_transit_count[transit])))
plt.xlabel('Time (BKJD, days)')
plt.ylabel('Flux')
# Test statistic
G = gridspec.GridSpec(9, 3)
G.update(hspace=0)
#G.update(hspace=3)
axes_7 = plt.subplot(G[6, :])
axes_7.axvline(results.period, alpha=0.4, lw=3)
for n in range(2, 5):
axes_7.axvline(n * results.period, alpha=0.4, lw=1, linestyle="dashed")
axes_7.axvline(results.period / n, alpha=0.4, lw=1, linestyle="dashed")
plt.ylabel(r'TLS SDE')
plt.xticks(())
plt.plot(results.periods, results.power, color='black', lw=0.5)
plt.xlim(0, max(results.periods))
plt.text(results.period + 1, results.SDE * 0.9,
'SDE=' + format(results.SDE, '.1f'))
# LS periodogram raw
#freqs, lspower = LombScargle(rawtime, rawflux).autopower()
freqs = numpy.geomspace(1 / min(results.periods), 1 / max(results.periods),
10000)
freqs = numpy.unique(freqs)
#print(freqs)
lspower = LombScargle(rawtime, rawflux).power(freqs)
freqs = 1 / freqs
#for i in range(len(freqs)):
# print(freqs[i], lspower[i])
#print('max(results.periods)', max(results.periods))
#idx_end = numpy.argmax(freqs<max(results.periods))
#print(lspower)
#print('idx_end', idx_end)
#print('idx_end2', numpy.argmax(freqs<5))
#freqs = freqs[:idx_end]
#lspower = lspower[:idx_end]
#print(lspower)
#continuum = numpy.std(lspower)
#lspower = numpy.divide(lspower, continuum) # Normalize to S/N
#print(lspower)
G = gridspec.GridSpec(9, 3)
G.update(hspace=0)
#G.update(hspace=0)
axes_8 = plt.subplot(G[7, :])
axes_8.axvline(results.period, alpha=0.4, lw=15)
#axes_8.set_yscale("log")
plt.ylabel(r'LS (raw)')
plt.plot(freqs, lspower, color='black', lw=0.5)
plt.xlim(0, max(results.periods))
plt.ylim(0, max(lspower) * 1.1)
plt.xlabel('Period (days)')
#idx_start = numpy.argmax(freqs>results.period-2)
#idx_end = numpy.argmax(freqs>results.period+2)
"""
lspower_at_period = max(lspower[idx_start:idx_end])
plt.text(results.period+1, max(lspower)*0.9, 'LS power=' + format(lspower_at_period, '.1f'))
if lspower_at_period > results.SDE:
text = 'sinusoidal model preferred'
else:
text = 'transit model preferred'
plt.text(results.period+1, max(lspower)*0.8, text)
"""
# Calculate Lomb-Scargle periodogram detrended
freqs = numpy.geomspace(1 / min(results.periods), 1 / max(results.periods),
10000)
freqs = numpy.unique(freqs)
lspower = LombScargle(t, y_filt).power(freqs)
freqs = 1 / freqs
#freqs, lspower = LombScargle(t, y_filt).autopower()
#idx_end = numpy.argmax(freqs>max(results.periods))
#freqs = freqs[:idx_end]
#lspower = lspower[:idx_end]
continuum = numpy.std(lspower)
lspower = numpy.divide(lspower, continuum) # Normalize to S/N
G = gridspec.GridSpec(9, 3)
G.update(hspace=0)
#G.update(hspace=3)
axes_9 = plt.subplot(G[8, :])
peak_index = numpy.argmax(lspower)
#print(peak_index, freqs[peak_index])
axes_9.axvline(freqs[peak_index], alpha=0.4, lw=10)
plt.ylabel(r'LS (detrended)')
plt.plot(freqs, lspower, color='black', lw=0.5)
plt.xlim(0, max(results.periods))
plt.ylim(0, max(lspower) * 1.1)
plt.xlabel('Period (days)')
#plt.text(freqs[peak_index]+1, max(lspower)*0.9, 'LS power=' + format(max(lspower), '.1f'))
#if max(lspower) > results.SDE:
# text = 'sinusoidal model preferred'
# model_preferred = 'sine'
#else:
# text = 'transit model preferred'
# model_preferred = 'transit'
#plt.text(results.period+1, max(lspower)*0.8, text)
# Text
G = gridspec.GridSpec(9, 3)
G.update(hspace=0, wspace=0.1)
axes_8 = plt.subplot(G[0:2, 2])
plt.xticks(())
plt.yticks(())
plt.text(0,
0,
'KOI ' + str(KOI) + '.0' + str(planet_number),
fontweight='bold')
#from astroquery.vizier import Vizier
#result = Vizier(columns=["Kpmag", "Dist"]).query_constraints(ID=EPIC_id, catalog="IV/34/epic")[0].as_array()
#Kpmag = result[0][0]
#Dist = result[0][1]
#plt.text(0, -0.15, 'KPmag=' + format(Kpmag, '.1f') + ', d=' + format(Dist, '.0f') + ' pc')
plt.text(
0, -0.30, 'SDE=' + format(results.SDE, '.1f') + ', SNR=' +
format(results.snr, '.1f'))
plt.text(
0, -0.45, 'P=' + format(results.period, '.5f') + ' +-' +
format(results.period_uncertainty, '.5f') + r'$\,$d')
plt.text(0, -0.60,
r'$T_{dur}=$' + format(results.duration, '.5f') + r'$\,$d')
plt.text(
0, -0.75,
r'$R_*=$' + format(radius, '.2f') + ' (+' + format(radius_max, '.2f') +
', -' + format(radius_min, '.2f') + ') $\,R_{\odot}$')
plt.text(
0, -0.90,
r'$M_*=$' + format(mass, '.2f') + ' (+' + format(mass_max, '.2f') +
', -' + format(mass_min, '.2f') + ') $\,M_{\odot}$')
plt.text(0, -1.05, r'$R_P/R_*=$' + format(results.rp_rs, '.3f'))
print('rp_rs', results.rp_rs)
rp = results.rp_rs * radius # in solar radii
sun = 695700
earth = 6371
jupi = 69911
rp_earth = (sun / earth) * rp
rp_jupi = (sun / jupi) * rp
plt.text(
0, -1.20, r'$R_P=$' + format(rp_earth, '.2f') + '$\,R_{\oplus}=$' +
format(rp_jupi, '.2f') + '$\,R_{Jup}$')
plt.text(
0, -1.35,
r'$\delta=$' + \
format((1-results.depth_mean[0])*10**6, '.0f') + ' ppm (' + \
format((1-results.depth_mean_odd[0])*10**6, '.0f') + ', ' + \
format((1-results.depth_mean_even[0])*10**6, '.0f') + ')')
plt.text(
0, -1.50, r'odd/even mismatch: ' +
format(results.odd_even_mismatch, '.2f') + '$\,\sigma$')
plt.text(
0, -1.65,
str(results.distinct_transit_count) + '/' +
str(results.transit_count) + ' transits with data')
plt.xlim(-0.1, 2)
plt.ylim(-2, 0.1)
axes_8.axis('off')
plt.subplots_adjust(hspace=0.4)
# Additional vetting criteria
valid = True
#if largest_phase_peak > 10:
# valid = False
# print('Vetting fail! largest_phase_peak > 10 median cadences: ', largest_phase_peak)
if abs(significance_eb) > 3:
valid = False
print('Vetting fail! significance_eb > 3')
#if numpy.isnan(results.odd_even_mismatch):
# valid = False
# print('Vetting fail! numpy.isnan(results.odd_even_mismatch)')
if abs(results.odd_even_mismatch) > 3:
valid = False
print('Vetting fail! odd_even_mismatch larger 3 sigma')
if transit_near_gaps and results.distinct_transit_count < 4:
valid = False
print('Vetting fail! Transit near gaps and distinct_transit_count < 4')
#if model_preferred != 'transit':
# valid = False
# print('Sinusoidal model was preferred over transit model')
#if max(lspower) > results.SDE:
# valid = False
# print('Vetting fail! Sinusoidal model preferred')
#if transit_touches_edge:
# valid = False
# print('Vetting fail! First or last transit touches edge of time series')
#valid = True
if valid:
figure_out_path = str(KOI) + '_0' + str(planet_number)
plt.savefig(figure_out_path + '.png', bbox_inches='tight')
#plt.savefig(figure_out_path + '.pdf', bbox_inches='tight')
print('Figure made:', figure_out_path)
"""
print('Creating full CSV result file')
numpy.savetxt(
fname=str(EPIC_id) + '_statistics.csv',
X=numpy.array([EPIC_id, results.period, results.T0]),
delimiter=',',
newline=" ",
fmt='%.5f',
)
"""
else:
print('Vetting criteria failed! No figure made.')
#plt.close()
print('T0', results.T0)
return valid