def test_kp_ks_k_conversion(makeSynthetic=False):
"""
For Arches photometry, we need to convert between Kp, Ks, and K
filters for extinction corrections and model comparisons. This
code tests whether we can simply use a linear relationship between
H-Kp and Kp-Ks or Kp-K for the conversion. I use synthetic atmospheres
for a simulated stellar population at 8 kpc, 2.5 Myr, solar metallicity,
for extinctions between AKs of 2-4 to get the range of colors.
"""
dir = '/u/jlu/work/arches/photo_calib/filter_conversions/'
os.chdir(dir)
if makeSynthetic:
synthetic.nearIR(8000, 6.4)
# Load up the synthetic data generated at a distance of
# 8 kpc and for a population with an age of 10**9.35 years.
synFile = dir + 'syn_nir_d08000_a640.dat'
T, AKs, J, H, K, Kp, Ks, Lp, mass, logg, logL = synthetic.load_nearIR(synFile)
# From Schoedel et al. 2010, the range of extinctions is given by a
# gaussian with mean AKs = 2.74 and stddev = 0.3. Our synthetic
# photometry samples AKs in steps of 0.1 so we will take the range
# 2.1 <= AKs <= 4.2 (see Espinoza et al. 2009).
adx = np.where((AKs >= 2.1) & (AKs <= 4.2))[0]
AKs = AKs[adx]
J = J[:,adx]
H = H[:,adx]
K = K[:,adx]
Kp = Kp[:,adx]
Ks = Ks[:,adx]
Lp = Lp[:,adx]
mass = mass[:,adx]
logg = logg[:,adx]
logL = logL[:,adx]
# Lets make a color scale for our range of AKs for plotting.
colorNorm = matplotlib.colors.Normalize(AKs)
# First lets plot the full range of temperatures and magnitudes
# to show what matches to the calibrators at Ks. Recall that our
# photometric calibrators have H < 16.5, Ks < 14.5, Lp < 13.5.
py.clf()
pltH = py.semilogx(T, H, color='b', label='H')
pltK = py.semilogx(T, Kp, color='g', label='Ks')
pltL = py.semilogx(T, Lp, color='r', label='Lp')
py.legend((pltH[0], pltK[0], pltL[0]), ('H', 'Kp', 'Lp'))
loc = matplotlib.ticker.MultipleLocator(10000)
py.gca().xaxis.set_major_locator( loc )
title_DA = 'd=8 kpc, 2.1<=AKs<=4.2'
py.xlim(40000, 3000)
py.ylim(25, 8)
py.xlabel('Effective Temperature (K)')
py.ylabel('Magnitude')
py.title(title_DA)
py.savefig(dir + 'temp_vs_mag.png')
##########
# Temperature range for calibrators (inclusive)
##########
T_range = [6000, 35000]
title_AT = 'AKs=[2.1:4.2] T=[6000:35000]'
# Lets trim down
tdx = np.where((T >= T_range[0]) & (T <= T_range[1]))[0]
T = T[tdx]
J = J[tdx,:]
H = H[tdx,:]
K = K[tdx,:]
Kp = Kp[tdx,:]
Ks = Ks[tdx,:]
Lp = Lp[tdx,:]
# Plot up the color-magnitude diagrams and compare with
# Schoedel et al. (2010).
AKs_2D = AKs.repeat(len(T)).reshape((len(AKs), len(T))).transpose()
# H-Ks vs. Ks
py.clf()
py.scatter(H-Ks, Ks, c=AKs_2D.flatten(), cmap=py.cm.jet, edgecolor='none')
cbar = py.colorbar(orientation='vertical', fraction=0.2)
cbar.set_label('AKs')
rng = py.axis()
py.xlim(0, 4)
py.ylim(25, 8)
py.xlabel('H - Ks')
py.ylabel('Ks')
py.title(title_AT)
py.savefig(dir + 'cmd_h_ks.png')
# Ks-Lp vs. Lp
py.clf()
py.scatter(Ks-Lp, Lp, c=AKs_2D.flatten(), cmap=py.cm.jet, edgecolor='none')
cbar = py.colorbar(orientation='vertical', fraction=0.2)
cbar.set_label('AKs')
rng = py.axis()
py.xlim(0, 3)
py.ylim(25, 8)
py.xlabel('Ks - Lp')
py.ylabel('Lp')
py.title(title_AT)
py.savefig(dir + 'cmd_ks_lp.png')
# Plot up the relationship between H-Ks and Kp-Ks.. same for Ks-Lp.
# In principle, we can use this relationship directly as long as any
# scatter for this simulated stellar population is less than 1%
# First, Lets fit a line to each relation
hkp = (H - Kp).flatten()
kpks = (Kp - Ks).flatten()
kplp = (Kp - Lp).flatten()
hkp_coeffs = np.polyfit(hkp, kpks, 1)
kplp_coeffs = np.polyfit(kplp, kpks, 1)
hkp_idx = hkp.argsort()
kplp_idx = kplp.argsort()
hkp_fit = np.polyval(hkp_coeffs, hkp[hkp_idx])
kplp_fit = np.polyval(kplp_coeffs, kplp[kplp_idx])
# H-Ks vs. Kp-Ks
py.clf()
py.scatter(H-Kp, Kp-Ks, c=AKs_2D.flatten(), cmap=py.cm.jet,
edgecolor='none')
py.plot(hkp[hkp_idx], hkp_fit, 'k--')
cbar = py.colorbar(orientation='vertical', fraction=0.2)
cbar.set_label('AKs')
py.xlabel('H - Kp')
py.ylabel('Kp - Ks')
py.title('Kp-Ks = %.4f + %.4f * H-Kp' % (hkp_coeffs[0], hkp_coeffs[1]),
fontsize=14)
py.savefig(dir+'color_HKp_KpKs.png')
diff = hkp_fit - kpks[hkp_idx]
print 'Best fit Kp-Ks = %.5f + %.5f * H-Ks' % \
(hkp_coeffs[0], hkp_coeffs[1])
print 'Residuals/Range from H-Kp vs. Kp-Ks fit: %.4f [%.4f - %.4f]' % \
(diff.std(), diff.min(), diff.max())
print ''
# Kp-Lp vs. Kp-Ks
py.clf()
py.scatter(Kp-Lp, Kp-Ks, c=AKs_2D.flatten(), cmap=py.cm.jet,
edgecolor='none')
py.plot(kplp[kplp_idx], kplp_fit, 'k--')
cbar = py.colorbar(orientation='vertical', fraction=0.2)
cbar.set_label('AKs')
py.xlabel('Kp - Lp')
py.ylabel('Kp - Ks')
py.title('Kp-Ks = %.4f + %.4f * Kp-Lp' % (kplp_coeffs[0], kplp_coeffs[1]),
fontsize=14)
py.savefig(dir+'color_KpLp_KpKs.png')
diff = kplp_fit - kpks[kplp_idx]
print 'Best fit Kp-Ks = %.5f + %.5f * Kp-Lp' % \
(kplp_coeffs[0], kplp_coeffs[1])
print 'Residuals/Range from Kp-Lp vs. Kp-Ks fit: %.4f [%.4f - %.4f]' % \
(diff.std(), diff.min(), diff.max())
print ''
# Lets do the same for Kp vs. K
hkp = (H - Kp).flatten()
kpk = (Kp - K).flatten()
kplp = (Kp - Lp).flatten()
hkp_coeffs = np.polyfit(hkp, kpk, 1)
kplp_coeffs = np.polyfit(kplp, kpk, 1)
hkp_idx = hkp.argsort()
kplp_idx = kplp.argsort()
hkp_fit = np.polyval(hkp_coeffs, hkp[hkp_idx])
kplp_fit = np.polyval(kplp_coeffs, kplp[kplp_idx])
# H-Kp vs. Kp-K
py.clf()
py.scatter(H-Kp, Kp-K, c=AKs_2D.flatten(), cmap=py.cm.jet,
edgecolor='none')
py.plot(hkp[hkp_idx], hkp_fit, 'k--')
cbar = py.colorbar(orientation='vertical', fraction=0.2)
cbar.set_label('AKs')
py.xlabel('H - Kp')
py.ylabel('Kp - K')
py.title('Kp-K = %.4f + %.4f * H-Kp' % (hkp_coeffs[0], hkp_coeffs[1]),
fontsize=14)
py.savefig(dir+'color_HKp_KpK.png')
diff = hkp_fit - kpk[hkp_idx]
print 'Best fit Kp-K = %.5f + %.5f * H-Kp' % \
(hkp_coeffs[0], hkp_coeffs[1])
print 'Residuals/Range from H-Kp vs. Kp-K fit: %.4f [%.4f - %.4f]' % \
(diff.std(), diff.min(), diff.max())
print ''
# Kp-Lp vs. Kp-K
py.clf()
py.scatter(Kp-Lp, Kp-K, c=AKs_2D.flatten(), cmap=py.cm.jet,
edgecolor='none')
py.plot(kplp[kplp_idx], kplp_fit, 'k--')
cbar = py.colorbar(orientation='vertical', fraction=0.2)
cbar.set_label('AKs')
py.xlabel('Kp - Lp')
py.ylabel('Kp - K')
py.title('Kp-K = %.4f + %.4f * Kp-Lp' % (kplp_coeffs[0], kplp_coeffs[1]),
fontsize=14)
py.savefig(dir+'color_KpLp_KpK.png')
diff = kplp_fit - kpk[kplp_idx]
print 'Best fit Kp-K = %.5f + %.5f * Kp-Lp' % \
(kplp_coeffs[0], kplp_coeffs[1])
print 'Residuals/Range from Kp-Lp vs. Kp-K fit: %.4f [%.4f - %.4f]' % \
(diff.std(), diff.min(), diff.max())
def make_synthetic():
synthetic.nearIR(distance,
logAge,
redlawClass=synthetic.RedLawRomanZuniga07)
def plot_cluster_isochrones(redo_iso=False):
"""
Plot isochrones and mass-luminosity functions for M17, Wd 2, Wd 1, and RSGC 1.
"""
# Cluster Info
name = ['M17', 'Wd 2', 'Wd 1', 'RSGC 1', 'RSGC 2']
dist = np.array([2100, 4160, 3600, 6000, 6000])
age = np.array([1., 2., 5., 12., 17.]) * 1.0e6
AV = np.array([5., 6.5, 10., 23., 10.])
# Derived properties
logage = np.log10(age)
AKs = AV / 10.0
iso_all = []
# Loop through the clusters and make the isochrones.
for ii in range(len(name)):
pickleFile = 'syn_nir_d' + str(dist[ii]).zfill(5) + '_a' \
+ str(int(round(logage[ii]*100))).zfill(3) + '.dat'
if (not os.path.exists(pickleFile)) or (redo_iso == True):
AKsGrid = np.array([AKs[ii]])
syn.nearIR(dist[ii], logage[ii], AKsGrid=AKsGrid)
iso_all.append( syn.load_nearIR_dict(pickleFile) )
##########
# Plot CMDs
##########
py.figure(1)
py.clf()
py.subplots_adjust(left=0.15)
colors = ['green', 'cyan', 'blue', 'red', 'purple']
for ii in range(len(iso_all)):
iso = iso_all[ii]
py.plot(iso['J'] - iso['K'], iso['K'], label=name[ii],
linewidth=2, color=colors[ii])
idx1 = np.argmin( np.abs(iso['mass'] - 1.0) )
py.plot(iso['J'][idx1] - iso['K'][idx1], iso['K'][idx1], 'ks',
color=colors[ii], mew=0, ms=8)
py.gca().invert_yaxis()
py.xlabel("J - K color")
py.ylabel("K magnitude")
py.text(iso['J'][idx1] - iso['K'][idx1], iso['K'][idx1],
r'1 M$_\odot$', color=colors[ii],
horizontalalignment='right', verticalalignment='top')
py.legend(loc="lower left")
py.savefig('clusters_cmd_jk.png')
##########
# Plot mass-luminosity relations for each of the filters.
##########
py.figure(2, figsize=(12,4))
py.clf()
py.subplots_adjust(left=0.06, bottom=0.15, wspace=0.22, right=0.97)
py.subplot(1, 3, 1)
for ii in range(len(iso_all)):
iso = iso_all[ii]
py.plot(iso['mass'], iso['J'], linewidth=2,
color=colors[ii], label=name[ii])
py.xlabel(r'Stellar Mass (M$_\dot$)')
py.ylabel('J magnitude')
py.xlim(0, 20)
py.ylim(26, 9)
py.subplot(1, 3, 2)
for ii in range(len(iso_all)):
iso = iso_all[ii]
py.plot(iso['mass'], iso['H'], linewidth=2,
color=colors[ii], label=name[ii])
py.legend(mode="expand", bbox_to_anchor=(-0.5, 0.99, 2.0, 0.08),
loc=3, ncol=4, frameon=False)
py.xlabel(r'Stellar Mass (M$_\dot$)')
py.ylabel('H magnitude')
py.xlim(0, 20)
py.ylim(26, 9)
py.subplot(1, 3, 3)
for ii in range(len(iso_all)):
iso = iso_all[ii]
py.plot(iso['mass'], iso['K'], linewidth=2,
color=colors[ii], label=name[ii])
py.xlabel(r'Stellar Mass (M$_\dot$)')
py.ylabel('K magnitude')
py.xlim(0, 20)
py.ylim(26, 9)
py.savefig('clusters_mass_luminosity_jhk.png')
##########
# Print out 0.1 Msun, 1 Msun, 10 Msun
# photometry table.
##########
for ii in range(len(iso_all)):
iso = iso_all[ii]
print('')
print('Cluster: ' + name[ii])
print('')
hdr = '{0:6s} {1:6s} {2:6s} {3:6s}'
dat = '{0:6.1f} {1:6.2f} {2:6.2f} {3:6.2f}'
print(hdr.format(' Mass', ' J', ' H', ' K'))
print(hdr.format(' ----', ' ---', ' ---', ' ---'))
idx01 = np.argmin( np.abs(iso['mass'] - 0.1) )
print(dat.format(iso['mass'][idx01], iso['J'][idx01][0],
iso['H'][idx01][0], iso['K'][idx01][0]))
idx1 = np.argmin( np.abs(iso['mass'] - 1.0) )
print(dat.format(iso['mass'][idx1], iso['J'][idx1][0],
iso['H'][idx1][0], iso['K'][idx1][0]))
idx10 = np.argmin( np.abs(iso['mass'] - 10.0) )
print(dat.format(iso['mass'][idx10], iso['J'][idx10][0],
iso['H'][idx10][0], iso['K'][idx10][0]))
def make_synthetic():
synthetic.nearIR(distance, logAge,
redlawClass=synthetic.RedLawRomanZuniga07)