def calc (inputDataDict, outfile, indir="simha/", lib="basel") :
from scipy.stats import chi2
import CosmologicalDistance
import os
cd = CosmologicalDistance.CosmologicalDistance()
ldistDict = dict()
splineDict = dict()
splines = loadPopColors.doall(indir, lib=lib)
id = inputDataDict["id"]
i = inputDataDict["i"]
ierr = inputDataDict["ierr"]
gr = inputDataDict["gr"]
ri = inputDataDict["ri"]
iz = inputDataDict["iz"]
grerr = inputDataDict["grerr"]
rierr = inputDataDict["rierr"]
izerr = inputDataDict["izerr"]
allzed = inputDataDict["zed"]
# protect against too small of errors => values = 0
ix = np.nonzero(grerr < 0.02)
grerr[ix] = 0.02
ix = np.nonzero(rierr < 0.02)
rierr[ix] = 0.02
ix = np.nonzero(izerr < 0.02)
izerr[ix] = 0.02
# prepping for output
out_id, out_gr, out_stdgr, out_gi, out_stdgi, \
out_kri, out_stdkri, out_kii, out_stdkii, out_iobs, out_distmod, \
out_rabs, out_iabs, out_mass_gr, out_mass_gi, out_mass, out_stdmass = \
[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[]
out_bestsp = []
size = id.size
#size = 10
for galaxy in range(0,size) :
zed = allzed[galaxy]
print galaxy," of ",size," z = ",zed
rest_gr,rest_gi,weight = [],[],[]
masslight, kii, kri = [],[],[]
minChiSq = 999; spIndex = -1
for sp in range(0,len(splines)) :
# for speed
skey = str(sp) + "-" + str(zed)
if skey in splineDict :
sgr,sri,siz,sgrr,sgir,skii,skri,sml = splineDict[skey]
else :
sgr = splines[sp][0](zed)
sri = splines[sp][1](zed)
siz = splines[sp][2](zed)
sgrr = splines[sp][4](zed) ;# restframe g-r
sgir = splines[sp][5](zed) ;# restframe g-i
skii = splines[sp][6](zed) ;# kcorrection: i_o - i_obs
skri = splines[sp][7](zed) ;# kcorrection: r_o - i_obs
sml = splines[sp][8](zed) ;# log(mass/light) (M_sun/L_sun)
splineDict[skey] = sgr,sri,siz,sgrr,sgir,skii,skri,sml
gre = grerr[galaxy]
rie = rierr[galaxy]
ize = izerr[galaxy]
gr_chisq = pow((gr[galaxy] - sgr)/gre,2)
ri_chisq = pow((ri[galaxy] - sri)/rie,2)
iz_chisq = pow((iz[galaxy] - siz)/ize,2)
rest_gr.append(sgrr)
rest_gi.append(sgir)
kii.append(skii)
kri.append(skri)
masslight.append(sml)
chisq = gr_chisq + ri_chisq + iz_chisq
probability = 1-chi2.cdf(chisq, 3-1) ;# probability of chisq greater than this
weight.append(probability)
spIndex = np.argmax(weight)
rest_gr = np.array(rest_gr)
rest_gi = np.array(rest_gi)
kii = np.array(kii)
kri = np.array(kri)
masslight = np.array(masslight)
weight = np.array(weight)
gr_weighted = rest_gr * weight
gi_weighted = rest_gi * weight
kii_weighted = kii * weight
kri_weighted = kri * weight
masslight_weighted = masslight * weight
w1 = weight.sum()
w2 = (weight**2).sum()
if w1 == 0 : w1 = 1e-10
if w2 == 0 : w2 = 1e-10
mean_gr = gr_weighted.sum()/w1
mean_gi = gi_weighted.sum()/w1
mean_kii = kii_weighted.sum()/w1
mean_kri = kri_weighted.sum()/w1
mean_masslight = masslight_weighted.sum()/w1
# unbiased weighted estimator of the sample variance
w3 = w1**2 - w2
if w3 == 0 : w3 = 1e-10
var_gr = ( w1/w3 ) * (weight*(rest_gr - mean_gr)**2).sum()
var_gi = ( w1/w3 ) * (weight*(rest_gi - mean_gi)**2).sum()
var_kii = ( w1/w3 ) * (weight*(kii - mean_kii)**2).sum()
var_kri = ( w1/w3 ) * (weight*(kri - mean_kii)**2).sum()
var_masslight = ( w1/w3 ) * (weight*(masslight - mean_masslight)**2).sum()
std_gr = var_gr**0.5
std_gi = var_gi**0.5
std_kii = var_kii**0.5
std_kri = var_kri**0.5
std_masslight = var_masslight**0.5
if std_gr > 99.99 : std_gr = 99.99
if std_gi > 99.99 : std_gi = 99.99
if std_kii > 99.99 : std_kii = 99.99
if std_kri > 99.99 : std_kri = 99.99
if std_masslight > 99.99 : std_masslight = 99.99
# Uncomment for fsps versions <2.5, as their mags don't include distance modulus
if zed in ldistDict :
lumdist = ldistDict[zed]
else :
lumdist = cd.luminosity_distance(zed) ;# in Mpc
ldistDict[zed] = lumdist
distanceModulus = 5*np.log10(lumdist*1e6/10.)
iabs = i[galaxy] + mean_kii - distanceModulus
rabs = i[galaxy] + mean_kri - distanceModulus
taMass = taylorMass(mean_gi, iabs)
mcMass = mcintoshMass(mean_gr, rabs)
fsMass = fspsMass( mean_masslight, iabs )
# JTA: to make purely distance modulus
#iabs = i[galaxy] - distanceModulus
#fsMass = gstarMass( iabs )
# saving for output
out_id.append( id[galaxy] )
out_gr.append( mean_gr )
out_stdgr.append(std_gr )
out_gi.append( mean_gi )
out_stdgi.append( std_gi)
out_kii.append( mean_kii )
out_stdkii.append( std_kii )
out_kri.append( mean_kri )
out_stdkri .append( std_kri )
out_iobs.append( i[galaxy] )
out_distmod.append( distanceModulus )
out_iabs.append( iabs )
out_rabs.append( rabs )
out_mass_gr.append( mcMass )
out_mass_gi.append( taMass )
out_mass.append( fsMass )
out_stdmass.append( std_masslight )
out_bestsp.append( spIndex )
out_id = np.array(out_id).astype(int)
out_gr = np.array(out_gr)
print out_gr
out_stdgr = np.array(out_stdgr)
out_gi = np.array(out_gi)
out_stdgi = np.array(out_stdgi)
out_kii = np.array(out_kii)
out_stdkii = np.array(out_stdkii)
out_kri = np.array(out_kri)
out_stdkri = np.array(out_stdkri )
out_iobs = np.array(out_iobs)
out_distmod = np.array(out_distmod)
out_iabs = np.array(out_iabs)
out_rabs = np.array(out_rabs)
out_mass_gr = np.array(out_mass_gr)
out_mass_gi = np.array(out_mass_gi)
out_mass = np.array(out_mass)
out_stdmass = np.array(out_stdmass)
out_bestsp = np.array(out_bestsp)
data = np.array([out_id, out_gr, out_stdgr, out_gi, out_stdgi, \
out_kri, out_stdkri, out_kii, out_stdkii, out_iobs, out_distmod, \
out_rabs, out_iabs, out_mass_gr, out_mass_gi, out_mass, out_stdmass, out_bestsp])
header = "# id, gr_o, std, gi_o, std, kri, std, kii, std, i, distmod, "
header = header + "r_abs, i_abs, mcMass, taMass, mass, std, bestsp\n"
fd = open(outfile,"w")
fd.write(header)
fd.close()
np.savetxt(outfile+".dat", data.T, "%d,%6.3f,%6.4f,%6.3f,%6.4f,%6.3f,%6.4f,%6.3f,%6.4f,\
%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%6.4f,%d")
os.system("cat {} >> {}; rm {}".format(outfile+".dat", outfile, outfile+".dat"))
def calc(inputDataDict, outfile, indir="simha/", lib="basel"):
from scipy.stats import chi2
import CosmologicalDistance
import os
cd = CosmologicalDistance.CosmologicalDistance()
ldistDict = dict()
splineDict = dict()
splines = loadPopColors.doall(indir, lib=lib)
id = inputDataDict["id"]
i = inputDataDict["i"]
ierr = inputDataDict["ierr"]
gr = inputDataDict["gr"]
ri = inputDataDict["ri"]
iz = inputDataDict["iz"]
grerr = inputDataDict["grerr"]
rierr = inputDataDict["rierr"]
izerr = inputDataDict["izerr"]
allzed = inputDataDict["zed"]
# protect against too small of errors => values = 0
ix = np.nonzero(grerr < 0.02)
grerr[ix] = 0.02
ix = np.nonzero(rierr < 0.02)
rierr[ix] = 0.02
ix = np.nonzero(izerr < 0.02)
izerr[ix] = 0.02
# prepping for output
out_id, out_gr, out_stdgr, out_gi, out_stdgi, out_kri, out_stdkri, out_kii, out_stdkii, out_iobs, out_distmod, out_rabs, out_iabs, out_mass_gr, out_mass_gi, out_mass, out_stdmass = (
[],
[],
[],
[],
[],
[],
[],
[],
[],
[],
[],
[],
[],
[],
[],
[],
[],
)
out_bestsp = []
size = id.size
# size = 10
for galaxy in range(0, size):
zed = allzed[galaxy]
print galaxy, " of ", size, " z = ", zed
rest_gr, rest_gi, weight = [], [], []
masslight, kii, kri = [], [], []
minChiSq = 999
spIndex = -1
for sp in range(0, len(splines)):
# for speed
skey = str(sp) + "-" + str(zed)
if skey in splineDict:
sgr, sri, siz, sgrr, sgir, skii, skri, sml = splineDict[skey]
else:
sgr = splines[sp][1](zed)
sri = splines[sp][2](zed)
siz = splines[sp][3](zed)
sgrr = splines[sp][4](zed)
# restframe g-r
sgir = splines[sp][5](zed)
# restframe g-i
skii = splines[sp][6](zed)
# kcorrection: i_o - i_obs
skri = splines[sp][7](zed)
# kcorrection: r_o - i_obs
sml = splines[sp][8](zed)
# log(mass/light) (M_sun/L_sun)
splineDict[skey] = sgr, sri, siz, sgrr, sgir, skii, skri, sml
gre = grerr[galaxy]
rie = rierr[galaxy]
ize = izerr[galaxy]
gr_chisq = pow((gr[galaxy] - sgr) / gre, 2)
ri_chisq = pow((ri[galaxy] - sri) / rie, 2)
iz_chisq = pow((iz[galaxy] - siz) / ize, 2)
rest_gr.append(sgrr)
rest_gi.append(sgir)
kii.append(skii)
kri.append(skri)
masslight.append(sml)
chisq = gr_chisq + ri_chisq + iz_chisq
probability = 1 - chi2.cdf(chisq, 3 - 1)
# probability of chisq greater than this
weight.append(probability)
spIndex = np.argmax(weight)
rest_gr = np.array(rest_gr)
rest_gi = np.array(rest_gi)
kii = np.array(kii)
kri = np.array(kri)
masslight = np.array(masslight)
weight = np.array(weight)
gr_weighted = rest_gr * weight
gi_weighted = rest_gi * weight
kii_weighted = kii * weight
kri_weighted = kri * weight
masslight_weighted = masslight * weight
w1 = weight.sum()
w2 = (weight ** 2).sum()
if w1 == 0:
w1 = 1e-10
if w2 == 0:
w2 = 1e-10
mean_gr = gr_weighted.sum() / w1
mean_gi = gi_weighted.sum() / w1
mean_kii = kii_weighted.sum() / w1
mean_kri = kri_weighted.sum() / w1
mean_masslight = masslight_weighted.sum() / w1
# unbiased weighted estimator of the sample variance
w3 = w1 ** 2 - w2
if w3 == 0:
w3 = 1e-10
var_gr = (w1 / w3) * (weight * (rest_gr - mean_gr) ** 2).sum()
var_gi = (w1 / w3) * (weight * (rest_gi - mean_gi) ** 2).sum()
var_kii = (w1 / w3) * (weight * (kii - mean_kii) ** 2).sum()
var_kri = (w1 / w3) * (weight * (kri - mean_kii) ** 2).sum()
var_masslight = (w1 / w3) * (weight * (masslight - mean_masslight) ** 2).sum()
std_gr = var_gr ** 0.5
std_gi = var_gi ** 0.5
std_kii = var_kii ** 0.5
std_kri = var_kri ** 0.5
std_masslight = var_masslight ** 0.5
if std_gr > 99.99:
std_gr = 99.99
if std_gi > 99.99:
std_gi = 99.99
if std_kii > 99.99:
std_kii = 99.99
if std_kri > 99.99:
std_kri = 99.99
if std_masslight > 99.99:
std_masslight = 99.99
if zed in ldistDict:
lumdist = ldistDict[zed]
else:
lumdist = cd.luminosity_distance(zed)
# in Mpc
ldistDict[zed] = lumdist
distanceModulus = 5 * np.log10(lumdist * 1e6 / 10.0)
iabs = i[galaxy] - distanceModulus + mean_kii
rabs = i[galaxy] - distanceModulus + mean_kri
taMass = taylorMass(mean_gi, iabs)
mcMass = mcintoshMass(mean_gr, rabs)
fsMass = fspsMass(mean_masslight, iabs)
# JTA: to make purely distance modulus
# iabs = i[galaxy] - distanceModulus
# fsMass = gstarMass( iabs )
# saving for output
out_id.append(id[galaxy])
out_gr.append(mean_gr)
out_stdgr.append(std_gr)
out_gi.append(mean_gi)
out_stdgi.append(std_gi)
out_kii.append(mean_kii)
out_stdkii.append(std_kii)
out_kri.append(mean_kri)
out_stdkri.append(std_kri)
out_iobs.append(i[galaxy])
out_distmod.append(distanceModulus)
out_iabs.append(iabs)
out_rabs.append(rabs)
out_mass_gr.append(mcMass)
out_mass_gi.append(taMass)
out_mass.append(fsMass)
out_stdmass.append(std_masslight)
out_bestsp.append(spIndex)
out_id = np.array(out_id).astype(int)
out_gr = np.array(out_gr)
out_stdgr = np.array(out_stdgr)
out_gi = np.array(out_gi)
out_stdgi = np.array(out_stdgi)
out_kii = np.array(out_kii)
out_stdkii = np.array(out_stdkii)
out_kri = np.array(out_kri)
out_stdkri = np.array(out_stdkri)
out_iobs = np.array(out_iobs)
out_distmod = np.array(out_distmod)
out_iabs = np.array(out_iabs)
out_rabs = np.array(out_rabs)
out_mass_gr = np.array(out_mass_gr)
out_mass_gi = np.array(out_mass_gi)
out_mass = np.array(out_mass)
out_stdmass = np.array(out_stdmass)
out_bestsp = np.array(out_bestsp)
data = np.array(
[
out_id,
out_gr,
out_stdgr,
out_gi,
out_stdgi,
out_kri,
out_stdkri,
out_kii,
out_stdkii,
out_iobs,
out_distmod,
out_rabs,
out_iabs,
out_mass_gr,
out_mass_gi,
out_mass,
out_stdmass,
out_bestsp,
]
)
header = "# id, gr_o, std, gi_o, std, kri, std, kii, std, i, distmod, "
header = header + "r_abs, i_abs, mcMass, taMass, mass, std, bestsp\n"
fd = open(outfile, "w")
fd.write(header)
fd.close()
np.savetxt(
outfile + ".dat",
data.T,
"%d,%6.3f,%6.4f,%6.3f,%6.4f,%6.3f,%6.4f,%6.3f,%6.4f,\
%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%6.4f,%d",
)
os.system("cat {} >> {}; rm {}".format(outfile + ".dat", outfile, outfile + ".dat"))
def calc(inputDataDict, indir="simha/", lib="miles"):
from scipy.stats import chi2
import CosmologicalDistance
import os
cd = CosmologicalDistance.CosmologicalDistance()
ldistDict = dict()
splineDict = dict()
splines, zmet = loadPopColors.doall(indir, lib=lib)
id = inputDataDict["id"]
haloid = inputDataDict["haloid"]
i = inputDataDict["i"]
ierr = inputDataDict["ierr"]
gr = inputDataDict["gr"]
ri = inputDataDict["ri"]
iz = inputDataDict["iz"]
grerr = inputDataDict["grerr"]
rierr = inputDataDict["rierr"]
izerr = inputDataDict["izerr"]
allzed = inputDataDict["zed"]
GR_P_COLOR = inputDataDict["GR_P_COLOR"]
RI_P_COLOR = inputDataDict["RI_P_COLOR"]
IZ_P_COLOR = inputDataDict["IZ_P_COLOR"]
GR_P_MEMBER = inputDataDict["GR_P_MEMBER"]
RI_P_MEMBER = inputDataDict["RI_P_MEMBER"]
IZ_P_MEMBER = inputDataDict["IZ_P_MEMBER"]
DIST_TO_CENTER = inputDataDict["DIST_TO_CENTER"]
GRP_RED = inputDataDict["GRP_RED"]
GRP_BLUE = inputDataDict["GRP_BLUE"]
RIP_RED = inputDataDict["RIP_RED"]
RIP_BLUE = inputDataDict["RIP_BLUE"]
IZP_RED = inputDataDict["IZP_RED"]
IZP_BLUE = inputDataDict["IZP_BLUE"]
# protect against too small of errors => values = 0
ix = np.nonzero(grerr < 0.02)
grerr[ix] = 0.02
ix = np.nonzero(rierr < 0.02)
rierr[ix] = 0.02
ix = np.nonzero(izerr < 0.02)
izerr[ix] = 0.02
mod_tab = np.genfromtxt('simha_miles_Nov16_models_lookup.tab')
t_start = mod_tab[:, 2]
# prepping for output
out_id,out_haloid, out_gr, out_stdgr, out_gi, out_stdgi, \
out_kri, out_stdkri, out_kii, out_stdkii, out_iobs, out_distmod, \
out_bestzmet,out_stdzmet, out_rabs, out_iabs, out_mass_gr, out_mass_gi, out_mass, out_stdmass, out_sfr, out_stdsfr, out_age, out_stdage, \
out_GR_P_COLOR,out_RI_P_COLOR,out_IZ_P_COLOR,out_GR_P_MEMBER,out_RI_P_MEMBER,out_IZ_P_MEMBER,out_DIST_TO_CENTER, \
out_GRP_RED,out_GRP_BLUE,out_RIP_RED,out_RIP_BLUE,out_IZP_RED,out_IZP_BLUE, out_zmet, out_zed =\
[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[]
out_bestsp, out_bestchisq = [], []
size = id.size
bins = 10
colors = iter(cm.rainbow(np.linspace(0, 1, bins)))
plt.clf()
#size = 10
for z in np.linspace(0.1, 1., num=bins):
zed = z
print z #galaxy," of ",size," z = ",zed
rest_gr, rest_gi, gr, ri, iz = [], [], [], [], []
masslight, sfrs,ages,ages_cosmic,ages_cut,zmets,kii, kri, models = [],[],[],[],[],[],[],[],[]
minChiSq = 999
spIndex = -1
for sp in range(0, len(splines)):
# for speed
skey = str(sp) + "-" + str(zed)
#if skey in splineDict :
# sgr,sri,siz,sgrr,sgir,skii,skri,sml = splineDict[skey]
#else :
# sgr = splines[sp][0](zed)
# sri = splines[sp][1](zed)
# siz = splines[sp][2](zed)
# sgrr = splines[sp][4](zed) ;# restframe g-r
# sgir = splines[sp][5](zed) ;# restframe g-i
# skii = splines[sp][6](zed) ;# kcorrection: i_o - i_obs
# skri = splines[sp][7](zed) ;# kcorrection: r_o - i_obs
# sml = splines[sp][8](zed) ;# log(mass/light) (M_sun/L_sun)
# ssfr = splines[sp][9](zed)
# sage_cosmic = splines[sp][10](zed)
# sage = splines[sp][11](zed)
# zmet = splines[12]
sgr = splines[sp][0](zed)
sri = splines[sp][1](zed)
siz = splines[sp][2](zed)
sgrr = splines[sp][4](zed)
# restframe g-r
sgir = splines[sp][5](zed)
# restframe g-i
skii = splines[sp][6](zed)
# kcorrection: i_o - i_obs
skri = splines[sp][7](zed)
# kcorrection: r_o - i_obs
sml = splines[sp][8](zed)
# log(mass/light) (M_sun/L_sun)
ssfr = splines[sp][9](zed)
if (ssfr < -20.): ssfr = -20.
sage_cosmic = splines[sp][10](zed)
sage = splines[sp][11](zed)
szmet = zmet[sp]
#To be changed if SFH changes
#splineDict[skey] = sgr,sri,siz,sgrr,sgir,skii,skri,sml
#print sage_cosmic
gr.append(sgr)
ri.append(sri)
iz.append(siz)
models.append(sp)
rest_gr.append(sgrr)
rest_gi.append(sgir)
kii.append(skii)
kri.append(skri)
masslight.append(sml)
sfrs.append(ssfr)
ages.append(sage)
ages_cosmic.append(sage_cosmic)
zmets.append(szmet)
ages_cut.append(10.**(sage_cosmic - 9.) - t_start[sp])
#plt.scatter(masslight,sfrs,edgecolor='none',alpha=0.7)
#plt.savefig('ssfr_vs_ml_z'+str(zed)+'.png')
#plt.clf()
print ages_cut
ages_cut = np.array(ages_cut)
sfrs = np.array(sfrs)
ages_cosmic = np.array(ages_cosmic)
idx = (ages_cut > 0.1)
plt.scatter(sfrs[idx],
ages_cosmic[idx],
edgecolor='none',
alpha=0.7,
label=str(zed),
color=next(colors))
#plt.savefig('restgr_vs_ssfr_z'+str(zed)+'.png')
#plt.clf()
#plt.scatter(ri,gr,edgecolor='none',alpha=0.7)
#plt.savefig('gr_vs_ri_z'+str(zed)+'.png')
#plt.clf()
plt.legend()
plt.xlabel('ssfr')
plt.ylabel('age')
plt.savefig('age_vs_ssfr_new.png')
def calc (inputDataDict, outfile, indir="simha/", lib="miles") :
from scipy.stats import chi2
import CosmologicalDistance
import os
cd = CosmologicalDistance.CosmologicalDistance()
ldistDict = dict()
splineDict = dict()
splines, zmet = loadPopColors.doall(indir, lib=lib)
id = inputDataDict["id"]
haloid = inputDataDict["haloid"]
i = inputDataDict["i"]
ierr = inputDataDict["ierr"]
gr = inputDataDict["gr"]
ri = inputDataDict["ri"]
iz = inputDataDict["iz"]
grerr = inputDataDict["grerr"]
rierr = inputDataDict["rierr"]
izerr = inputDataDict["izerr"]
allzed = inputDataDict["zed"]
GR_P_COLOR=inputDataDict["GR_P_COLOR"]
RI_P_COLOR=inputDataDict["RI_P_COLOR"]
IZ_P_COLOR=inputDataDict["IZ_P_COLOR"]
P_RADIAL=inputDataDict["P_RADIAL"]
P_REDSHIFT=inputDataDict["P_REDSHIFT"]
P_MEMBER=inputDataDict["P_MEMBER"]
DIST_TO_CENTER=inputDataDict["DIST_TO_CENTER"]
GRP_RED=inputDataDict["GRP_RED"]
GRP_BLUE=inputDataDict["GRP_BLUE"]
RIP_RED=inputDataDict["RIP_RED"]
RIP_BLUE=inputDataDict["RIP_BLUE"]
IZP_RED=inputDataDict["IZP_RED"]
IZP_BLUE=inputDataDict["IZP_BLUE"]
#print zmet
# protect against too small of errors => values = 0
ix = np.nonzero(grerr < 0.02)
grerr[ix] = 0.02
ix = np.nonzero(rierr < 0.02)
rierr[ix] = 0.02
ix = np.nonzero(izerr < 0.02)
izerr[ix] = 0.02
# prepping for output
out_id,out_haloid, out_gr, out_stdgr, out_gi, out_stdgi, \
out_kri, out_stdkri, out_kii, out_stdkii, out_iobs, out_distmod, \
out_bestzmet,out_stdzmet, out_rabs, out_iabs, out_mass_gr, out_mass_gi, out_mass, out_stdmass, out_sfr, out_stdsfr, out_age, out_stdage, \
out_GR_P_COLOR,out_RI_P_COLOR,out_IZ_P_COLOR,out_P_RADIAL,out_P_REDSHIFT,out_P_MEMBER,out_DIST_TO_CENTER, \
out_GRP_RED,out_GRP_BLUE,out_RIP_RED,out_RIP_BLUE,out_IZP_RED,out_IZP_BLUE,out_zmet, out_zed =\
[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[]
out_bestsp, out_bestchisq = [], []
size = id.size
#size = 10
for galaxy in range(0,size) :
zed = allzed[galaxy]
print galaxy+1," of ",size," z = ",zed
rest_gr,rest_gi,weight,chisqs = [], [],[],[]
masslight, sfrs,ages, zmets,kii, kri = [],[],[],[],[],[]
minChiSq = 999; spIndex = -1
for sp in range(0,len(splines)) :
# for speed
skey = str(sp) + "-" + str(zed)
#if skey in splineDict :
# sgr,sri,siz,sgrr,sgir,skii,skri,sml = splineDict[skey]
#else :
# sgr = splines[sp][0](zed)
# sri = splines[sp][1](zed)
# siz = splines[sp][2](zed)
# sgrr = splines[sp][4](zed) ;# restframe g-r
# sgir = splines[sp][5](zed) ;# restframe g-i
# skii = splines[sp][6](zed) ;# kcorrection: i_o - i_obs
# skri = splines[sp][7](zed) ;# kcorrection: r_o - i_obs
# sml = splines[sp][8](zed) ;# log(mass/light) (M_sun/L_sun)
# ssfr = splines[sp][9](zed)
# sage_cosmic = splines[sp][10](zed)
# sage = splines[sp][11](zed)
# zmet = splines[12]
sgr = splines[sp][0](zed)
sri = splines[sp][1](zed)
siz = splines[sp][2](zed)
sgrr = splines[sp][4](zed) ;# restframe g-r
sgir = splines[sp][5](zed) ;# restframe g-i
skii = splines[sp][6](zed) ;# kcorrection: i_o - i_obs
skri = splines[sp][7](zed) ;# kcorrection: r_o - i_obs
sml = splines[sp][8](zed) ;# log(mass/light) (M_sun/L_sun)
ssfr = splines[sp][9](zed)
if (ssfr<-20.): ssfr=-20.
sage_cosmic = splines[sp][10](zed)
sage = splines[sp][11](zed)
szmet = zmet[sp]
#To be changed if SFH changes
#splineDict[skey] = sgr,sri,siz,sgrr,sgir,skii,skri,sml
gre = grerr[galaxy]
rie = rierr[galaxy]
ize = izerr[galaxy]
gr_chisq = pow((gr[galaxy] - sgr)/gre,2)
ri_chisq = pow((ri[galaxy] - sri)/rie,2)
iz_chisq = pow((iz[galaxy] - siz)/ize,2)
rest_gr.append(sgrr)
rest_gi.append(sgir)
kii.append(skii)
kri.append(skri)
masslight.append(sml)
sfrs.append(ssfr)
ages.append(sage)
zmets.append(szmet)
chisq = gr_chisq + ri_chisq + iz_chisq
probability = 1-chi2.cdf(chisq, 3-1) ;# probability of chisq greater than this
weight.append(probability)
chisqs.append(chisq)
spIndex = np.argmax(weight)
rest_gr = np.array(rest_gr)
rest_gi = np.array(rest_gi)
kii = np.array(kii)
kri = np.array(kri)
masslight = np.array(masslight)
sfrs = np.array(sfrs)
idx_sfr = (sfrs<-8.)
ages = np.array(ages)
weight = np.array(weight)
gr_weighted = rest_gr * weight
gi_weighted = rest_gi * weight
kii_weighted = kii * weight
kri_weighted = kri * weight
#weight_norm = weight/np.sum(weight)
masslight_weighted = masslight * weight
sfr_weighted = 10**sfrs[idx_sfr] * weight[idx_sfr]
age_weighted = ages * weight
zmet_weighted = zmets * weight
w1 = weight.sum()
w2 = (weight**2).sum()
if w1 == 0 : w1 = 1e-10
if w2 == 0 : w2 = 1e-10
mean_gr = gr_weighted.sum()/w1
mean_gi = gi_weighted.sum()/w1
mean_kii = kii_weighted.sum()/w1
mean_kri = kri_weighted.sum()/w1
mean_masslight = masslight_weighted.sum()/w1
#try :
# if weight.shape[0]>1.:
# mean_sfr = float(ws.numpy_weighted_median(sfrs, weights=weight)) #np.median(sfr_weighted) #sfr_weighted.sum()/w1
# else:
# mean_sfr = -70.
#except :
# mean_sfr = -70.
#print mean_sfr
mean_age = age_weighted.sum()/w1
mean_zmet = zmet_weighted.sum()/w1
mean_sfr = sfr_weighted.sum()/w1
# unbiased weighted estimator of the sample variance
w3 = w1**2 - w2
if w3 == 0 : w3 = 1e-10
var_gr = ( w1/w3 ) * (weight*(rest_gr - mean_gr)**2).sum()
var_gi = ( w1/w3 ) * (weight*(rest_gi - mean_gi)**2).sum()
var_kii = ( w1/w3 ) * (weight*(kii - mean_kii)**2).sum()
var_kri = ( w1/w3 ) * (weight*(kri - mean_kii)**2).sum()
var_masslight = ( w1/w3 ) * (weight*(masslight - mean_masslight)**2).sum()
var_sfr = ( w1/w3 ) * (weight*(sfrs - mean_sfr)**2).sum()
var_age = ( w1/w3 ) * (weight*(ages - mean_age)**2).sum()
var_zmet = ( w1/w3 ) * (weight*(zmets - mean_zmet)**2).sum()
std_gr = var_gr**0.5
std_gi = var_gi**0.5
std_kii = var_kii**0.5
std_kri = var_kri**0.5
std_masslight = var_masslight**0.5
std_sfr = var_sfr**0.5
std_age = var_age**0.5
std_zmet = var_zmet**0.5
if std_gr > 99.99 : std_gr = 99.99
if std_gi > 99.99 : std_gi = 99.99
if std_kii > 99.99 : std_kii = 99.99
if std_kri > 99.99 : std_kri = 99.99
if std_sfr > 99.99 : std_sfr = 99.99
if std_age > 99.99 : std_age = 99.99
if std_masslight > 99.99 : std_masslight = 99.99
if std_zmet > 99.99 : std_zmet = 99.99
# Comment -distanceModulus out for fsps versions <2.5, as their mags don't include distance modulus
if zed in ldistDict :
lumdist = ldistDict[zed]
else :
lumdist = cd.luminosity_distance(zed) ;# in Mpc
ldistDict[zed] = lumdist
distanceModulus = 5*np.log10(lumdist*1e6/10.)
iabs = i[galaxy] + mean_kii - distanceModulus
rabs = i[galaxy] + mean_kri - distanceModulus
taMass = taylorMass(mean_gi, iabs)
mcMass = mcintoshMass(mean_gr, rabs)
fsMass = fspsMass( mean_masslight, iabs )
# JTA: to make purely distance modulus
#iabs = i[galaxy] - distanceModulus
#fsMass = gstarMass( iabs )
# saving for output
out_id.append( id[galaxy] )
out_haloid.append( haloid[galaxy] )
out_gr.append( mean_gr )
out_stdgr.append(std_gr )
out_gi.append( mean_gi )
out_stdgi.append( std_gi)
out_kii.append( mean_kii )
out_stdkii.append( std_kii )
out_kri.append( mean_kri )
out_stdkri .append( std_kri )
out_iobs.append( i[galaxy] )
out_distmod.append( distanceModulus )
out_iabs.append( iabs )
out_rabs.append( rabs )
out_mass_gr.append( mcMass )
out_mass_gi.append( taMass )
out_mass.append( fsMass )
out_stdmass.append( std_masslight )
out_bestsp.append( spIndex )
out_bestzmet.append( zmets[spIndex] )
out_bestchisq.append(chisqs[spIndex])
out_sfr.append( mean_sfr )
out_stdsfr.append( std_sfr )
out_age.append( mean_age )
out_stdage.append( std_age )
out_zmet.append( mean_zmet )
out_stdzmet.append( std_zmet )
out_zed.append( allzed[galaxy] )
out_GR_P_COLOR.append( GR_P_COLOR[galaxy])
out_RI_P_COLOR.append( RI_P_COLOR[galaxy])
out_IZ_P_COLOR.append( IZ_P_COLOR[galaxy])
out_P_RADIAL.append( P_RADIAL[galaxy])
out_P_REDSHIFT.append( P_REDSHIFT[galaxy])
out_P_MEMBER.append( P_MEMBER[galaxy])
out_DIST_TO_CENTER.append( DIST_TO_CENTER[galaxy])
out_GRP_RED.append( GRP_RED[galaxy])
out_GRP_BLUE.append( GRP_BLUE[galaxy])
out_RIP_RED.append( RIP_RED[galaxy])
out_RIP_BLUE.append( RIP_BLUE[galaxy])
out_IZP_RED.append( IZP_RED[galaxy])
out_IZP_BLUE.append( IZP_BLUE[galaxy])
out_id = np.array(out_id).astype(int)
out_haloid = np.array(out_haloid).astype(int)
out_gr = np.array(out_gr)
out_stdgr = np.array(out_stdgr)
out_gi = np.array(out_gi)
out_stdgi = np.array(out_stdgi)
out_kii = np.array(out_kii)
out_stdkii = np.array(out_stdkii)
out_kri = np.array(out_kri)
out_stdkri = np.array(out_stdkri )
out_iobs = np.array(out_iobs)
out_distmod = np.array(out_distmod)
out_iabs = np.array(out_iabs)
out_rabs = np.array(out_rabs)
out_mass_gr = np.array(out_mass_gr)
out_mass_gi = np.array(out_mass_gi)
out_mass = np.array(out_mass)
out_stdmass = np.array(out_stdmass)
out_sfr = np.array(out_sfr)
out_stdsfr = np.array(out_stdsfr)
out_age = np.array(out_age)
out_stdage = np.array(out_stdage)
out_bestsp = np.array(out_bestsp)
out_GR_P_COLOR = np.array(out_GR_P_COLOR)
out_RI_P_COLOR = np.array(out_RI_P_COLOR)
out_IZ_P_COLOR = np.array(out_IZ_P_COLOR)
out_P_RADIAL = np.array(out_P_RADIAL)
out_P_REDSHIFT = np.array(out_P_REDSHIFT)
out_P_MEMBER = np.array(out_P_MEMBER)
out_DIST_TO_CENTER = np.array(out_DIST_TO_CENTER)
out_GRP_RED = np.array(out_GRP_RED)
out_GRP_BLUE = np.array(out_GRP_BLUE)
out_RIP_RED = np.array(out_RIP_RED)
out_RIP_BLUE = np.array(out_RIP_BLUE)
out_IZP_RED = np.array(out_IZP_RED)
out_IZP_BLUE = np.array(out_IZP_BLUE)
out_zmet = np.array(out_zmet)
out_bestzmet = np.array(out_bestzmet)
out_zed =np.array(out_zed)
out_bestchisq = np.array(out_bestchisq)
col1=pf.Column(name='MEM_MATCH_ID',format='J',array=out_haloid)
col2=pf.Column(name='Z',format='E',array=out_zed)
col3=pf.Column(name='ID',format='K',array=out_id)
col4=pf.Column(name='gr_o',format='E',array=out_gr)
col5=pf.Column(name='gr_o_err',format='E',array=out_stdgr)
col6=pf.Column(name='gi_o',format='E',array=out_gi)
col7=pf.Column(name='gi_o_err',format='E',array=out_stdgi)
col8=pf.Column(name='kri',format='E',array=out_kri)
col9=pf.Column(name='kri_err',format='E',array=out_stdkri)
col10=pf.Column(name='kii',format='E',array=out_kii)
col11=pf.Column(name='kii_err',format='E',array=out_stdkii)
col12=pf.Column(name='iobs',format='E',array=out_iobs)
col13=pf.Column(name='distmod',format='E',array=out_distmod)
col14=pf.Column(name='rabs',format='E',array=out_rabs)
col15=pf.Column(name='iabs',format='E',array=out_iabs)
col16=pf.Column(name='mcMass',format='E',array=out_mass_gr)
col17=pf.Column(name='taMass',format='E',array=out_mass_gi)
col18=pf.Column(name='mass',format='E',array=out_mass)
col19=pf.Column(name='mass_err',format='E',array=out_stdmass)
col20=pf.Column(name='ssfr',format='E',array=out_sfr)
col21=pf.Column(name='ssfr_std',format='E',array=out_stdsfr)
col22=pf.Column(name='mass_weight_age',format='E',array=out_age)
col23=pf.Column(name='mass_weight_age_err',format='E',array=out_stdage)
col24=pf.Column(name='best_model',format='E',array=out_bestsp)
col25=pf.Column(name='best_zmet',format='E',array=out_bestzmet)
col26=pf.Column(name='zmet',format='E',array=out_zmet)
col27=pf.Column(name='best_chisq',format='E',array=out_bestchisq)
col28=pf.Column(name='GR_P_COLOR',format='E',array=out_GR_P_COLOR)
col29=pf.Column(name='RI_P_COLOR',format='E',array=out_RI_P_COLOR)
col30=pf.Column(name='IZ_P_COLOR',format='E',array=out_IZ_P_COLOR)
col31=pf.Column(name='P_RADIAL',format='E',array=out_P_RADIAL)
col32=pf.Column(name='P_REDSHIFT',format='E',array=out_P_REDSHIFT)
col33=pf.Column(name='P_MEMBER',format='E',array=out_P_MEMBER)
col34=pf.Column(name='DIST_TO_CENTER',format='E',array=out_DIST_TO_CENTER)
col35=pf.Column(name='GRP_RED',format='E',array=out_GRP_RED)
col36=pf.Column(name='GRP_BLUE',format='E',array=out_GRP_BLUE)
col37=pf.Column(name='RIP_RED',format='E',array=out_RIP_RED)
col38=pf.Column(name='RIP_BLUE',format='E',array=out_RIP_BLUE)
col39=pf.Column(name='IZP_RED',format='E',array=out_IZP_RED)
col40=pf.Column(name='IZP_BLUE',format='E',array=out_IZP_BLUE)
cols=pf.ColDefs([col1,col2,col3,col4,col5,col6,col7,col8,col9,col10,col11,col12,col13,col14,col15,col16,col17,col18,col19,col20,col21,col22,col23,col24,col25,col26,col27,col28,col29,col30,col31,col32,col33,col34,col35,col36,col37,col38,col39,col40])
tbhdu=pf.BinTableHDU.from_columns(cols)
tbhdu.writeto(outfile,clobber=True)
print
print 'Exiting smass.py'
print
data = np.array([out_id, out_haloid, out_gr, out_stdgr, out_gi, out_stdgi, \
out_kri, out_stdkri, out_kii, out_stdkii, out_iobs, out_distmod, \
out_rabs, out_iabs, out_mass_gr, out_mass_gi, out_mass, out_stdmass, \
out_zed,out_sfr, out_stdsfr, out_age, out_stdage, out_bestsp,out_bestzmet,out_zmet,out_GR_P_COLOR,out_RI_P_COLOR,out_IZ_P_COLOR,out_P_RADIAL,out_P_REDSHIFT,out_P_MEMBER,out_DIST_TO_CENTER, \
out_GRP_RED,out_GRP_BLUE,out_RIP_RED,out_RIP_BLUE,out_IZP_RED,out_IZP_BLUE])
def calc(inputDataDict, outfile, indir="simha/", lib="miles"):
from scipy.stats import chi2
import CosmologicalDistance
import os
n_outputs = 26
Njobs = 3
cd = CosmologicalDistance.CosmologicalDistance()
ldistDict = dict()
splineDict = dict()
splines, zmet = loadPopColors.doall(indir, lib=lib)
id = inputDataDict["id"]
i = inputDataDict["i"]
ierr = inputDataDict["ierr"]
gr = inputDataDict["gr"]
ri = inputDataDict["ri"]
iz = inputDataDict["iz"]
grerr = inputDataDict["grerr"]
rierr = inputDataDict["rierr"]
izerr = inputDataDict["izerr"]
zed = inputDataDict["zed"]
size = i.size
#size=3
# protect against too small of errors => values = 0
ix = np.nonzero(grerr < 0.02)
grerr[ix] = 0.02
ix = np.nonzero(rierr < 0.02)
rierr[ix] = 0.02
ix = np.nonzero(izerr < 0.02)
izerr[ix] = 0.02
size = 10
#folder = tempfile.mkdtemp()
#outputs_name = os.path.join(folder, 'out')
# Pre-allocate a writeable shared memory map as a container for the
# results of the parallel computation
#outputs = np.memmap(outputs_name,dtype='float32',shape=((size,26)), mode='w+')
# put all inputs in one array, it's horrible this way...
outputs = Parallel(n_jobs=Njobs)(delayed(
gal_smass)(id[galaxy], i[galaxy], ierr[galaxy], gr[galaxy], ri[galaxy],
iz[galaxy], grerr[galaxy], rierr[galaxy], izerr[galaxy],
zed[galaxy], splines, zmet, galaxy)
for galaxy in range(0, size))
#print outputs
#print outputs.shape
header = "# id, photoz, gr_o, std, gi_o, std, kri, std, kii, std, i, distmod, "
header = header + "r_abs, i_abs, mcMass, taMass, mass, std, zed, sfr, sfrstd, age, agestd, bestsp,best_zmet,mean_zmet \n"
fd = open(outfile, "w")
fd.write(header)
fd.close()
np.savetxt(
outfile + ".dat",
outputs,
fmt=
"%d,%6.3f,%6.4f,%6.3f,%6.4f,%6.3f,%6.4f,%6.3f,%6.4f,%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%6.3f,%d,%6.3f,%6.4f,%6.3f,%6.3f,%6.3f,%6.3f, %6.3f"
)
os.system("cat {} >> {}; rm {}".format(outfile + ".dat", outfile,
outfile + ".dat"))
def nike(output, indir="simha/", vtcFile="../vtColors_annis.csv"):
from scipy.stats import chi2
import CosmologicalDistance
cd = CosmologicalDistance.CosmologicalDistance()
ldistDict = dict()
splineDict = dict()
splines = loadPopColors.doall(indir)
memberData = readVTmembers()
id, i, ug, gr, ri, iz, ugerr, grerr, rierr, izerr = readVT(vtcFile)
# protect against too small of errors => values = 0
ix = np.nonzero(ugerr < 0.02)
ugerr[ix] = 0.02
ix = np.nonzero(grerr < 0.02)
grerr[ix] = 0.02
ix = np.nonzero(rierr < 0.02)
rierr[ix] = 0.02
ix = np.nonzero(izerr < 0.02)
izerr[ix] = 0.02
gal_id = memberData[4]
gal_z = memberData[2]
# prepping for output
gal_ra = memberData[0]
gal_dec = memberData[1]
gal_ze = memberData[3]
gal_central = memberData[5]
gal_hostid = memberData[6]
# prepping for output
out_ra, out_dec, out_z, out_ze = [], [], [], []
out_id, out_central, out_hostid = [], [], []
out_gr, out_stdgr, out_gi, out_stdgi = [], [], [], []
out_kii, out_stdkii, out_kri, out_stdkri = [], [], [], []
out_iobs, out_distmod, out_iabs, out_rabs, out_mass_gr, out_mass_gi= [],[],[],[],[],[]
out_mass, out_stdmass = [], []
size = id.size
#size = 10
for galaxy in range(0, size):
ix = np.nonzero(gal_id == id[galaxy])
zed = gal_z[ix][0]
if zed == 0.0: continue
print galaxy, " of ", size
rest_gr, rest_gi, weight = [], [], []
masslight, kii, kri = [], [], []
for sp in range(0, len(splines)):
# for speed
skey = str(sp) + "-" + str(zed)
if skey in splineDict:
sug, sgr, sri, siz, sgrr, sgir, skii, skri, sml = splineDict[
skey]
else:
sug = splines[sp][0](zed)
sgr = splines[sp][1](zed)
sri = splines[sp][2](zed)
siz = splines[sp][3](zed)
sgrr = splines[sp][4](zed)
# restframe g-r
sgir = splines[sp][5](zed)
# restframe g-i
skii = splines[sp][6](zed)
# kcorrection: i_o - i_obs
skri = splines[sp][7](zed)
# kcorrection: r_o - i_obs
sml = splines[sp][8](zed)
# log(mass/light) (M_sun/L_sun)
splineDict[
skey] = sug, sgr, sri, siz, sgrr, sgir, skii, skri, sml
uge = ugerr[galaxy]
gre = grerr[galaxy]
rie = rierr[galaxy]
ize = izerr[galaxy]
ug_chisq = pow((ug[galaxy] - sug) / uge, 2)
gr_chisq = pow((gr[galaxy] - sgr) / gre, 2)
ri_chisq = pow((ri[galaxy] - sri) / rie, 2)
iz_chisq = pow((iz[galaxy] - siz) / ize, 2)
rest_gr.append(sgrr)
rest_gi.append(sgir)
kii.append(skii)
kri.append(skri)
masslight.append(sml)
chisq = ug_chisq + gr_chisq + ri_chisq + iz_chisq
probability = 1 - chi2.cdf(chisq, 3)
# probability of chisq greater than this
weight.append(probability)
rest_gr = np.array(rest_gr)
rest_gi = np.array(rest_gi)
kii = np.array(kii)
kri = np.array(kri)
masslight = np.array(masslight)
weight = np.array(weight)
gr_weighted = rest_gr * weight
gi_weighted = rest_gi * weight
kii_weighted = kii * weight
kri_weighted = kri * weight
masslight_weighted = masslight * weight
# weighted sum
w1 = weight.sum()
w2 = (weight**2).sum()
if w1 == 0: w1 = 1e-10
if w2 == 0: w2 = 1e-10
mean_gr = gr_weighted.sum() / w1
mean_gi = gi_weighted.sum() / w1
mean_kii = kii_weighted.sum() / w1
mean_kri = kri_weighted.sum() / w1
mean_masslight = masslight_weighted.sum() / w1
# unbiased weighted estimator of the sample variance
w3 = w1**2 - w2
if w3 == 0: w3 = 1e-10
var_gr = (w1 / w3) * (weight * (rest_gr - mean_gr)**2).sum()
var_gi = (w1 / w3) * (weight * (rest_gi - mean_gi)**2).sum()
var_kii = (w1 / w3) * (weight * (kii - mean_kii)**2).sum()
var_kri = (w1 / w3) * (weight * (kri - mean_kii)**2).sum()
var_masslight = (w1 / w3) * (weight *
(masslight - mean_masslight)**2).sum()
std_gr = var_gr**0.5
std_gi = var_gi**0.5
std_kii = var_kii**0.5
std_kri = var_kri**0.5
std_masslight = var_masslight**0.5
if std_gr > 99.99: std_gr = 99.99
if std_gi > 99.99: std_gi = 99.99
if std_kii > 99.99: std_kii = 99.99
if std_kri > 99.99: std_kri = 99.99
if std_masslight > 99.99: std_masslight = 99.99
if zed in ldistDict:
lumdist = ldistDict[zed]
else:
lumdist = cd.luminosity_distance(zed)
# in Mpc
ldistDict[zed] = lumdist
distanceModulus = 5 * np.log10(lumdist * 1e6 / 10.)
iabs = i[galaxy] - distanceModulus + mean_kii
rabs = i[galaxy] - distanceModulus + mean_kri
taMass = taylorMass(mean_gi, iabs)
mcMass = mcintoshMass(mean_gr, rabs)
fsMass = fspsMass(mean_masslight, iabs)
# saving for output
out_ra.append(gal_ra[ix][0])
out_dec.append(gal_dec[ix][0])
out_z.append(gal_z[ix][0])
out_ze.append(gal_ze[ix][0])
out_id.append(gal_id[ix][0])
out_central.append(gal_central[ix][0])
out_hostid.append(gal_hostid[ix][0])
out_gr.append(mean_gr)
out_stdgr.append(std_gr)
out_gi.append(mean_gi)
out_stdgi.append(std_gi)
out_kii.append(mean_kii)
out_stdkii.append(std_kii)
out_kri.append(mean_kri)
out_stdkri.append(std_kri)
out_iobs.append(i[galaxy])
out_distmod.append(distanceModulus)
out_iabs.append(iabs)
out_rabs.append(rabs)
out_mass_gr.append(mcMass)
out_mass_gi.append(taMass)
out_mass.append(fsMass)
out_stdmass.append(std_masslight)
out_ra = np.array(out_ra)
out_dec = np.array(out_dec)
out_z = np.array(out_z)
out_ze = np.array(out_ze)
out_id = np.array(out_id).astype(int)
out_central = np.array(out_central).astype(int)
out_hostid = np.array(out_hostid).astype(int)
out_gr = np.array(out_gr)
out_stdgr = np.array(out_stdgr)
out_gi = np.array(out_gi)
out_stdgi = np.array(out_stdgi)
out_kii = np.array(out_kii)
out_stdkii = np.array(out_stdkii)
out_kri = np.array(out_kri)
out_stdkri = np.array(out_stdkri)
out_iobs = np.array(out_iobs)
out_distmod = np.array(out_distmod)
out_iabs = np.array(out_iabs)
out_rabs = np.array(out_rabs)
out_mass_gr = np.array(out_mass_gr)
out_mass_gi = np.array(out_mass_gi)
out_mass = np.array(out_mass)
out_stdmass = np.array(out_stdmass)
data = np.array([out_ra, out_dec, out_z, out_ze, out_id, \
out_central, out_hostid, out_gr, out_stdgr, out_gi, out_stdgi, \
out_kri, out_stdkri, out_kii, out_stdkii, out_iobs, out_distmod, \
out_rabs, out_iabs, out_mass_gr, out_mass_gi, out_mass, out_stdmass])
np.savetxt(
output, data.T, "%10.6f %10.5f %4.2f %5.3f %5d %d %10d \
%6.3f %6.4f %6.3f %6.4f %6.3f %6.4f %6.3f %6.4f \
%6.3f %6.3f %6.3f %6.3f %6.3f %6.3f %6.3f %6.4f")
fd = open(output + ".hdr", "w")
fd.write(
"# ra dec z zerr id central hostid g-r_o std g-i_o "
)
fd.write(
"std kri std kii std i distmod r_abs i_abs mcMass taMass mass std\n"
)
fd.close()
def calc(inputDataDict, outfile, indir="simha/", lib="miles"):
from scipy.stats import chi2
import CosmologicalDistance
import os
cd = CosmologicalDistance.CosmologicalDistance()
ldistDict = dict()
splineDict = dict()
print "Loading " + lib + " library..."
splines, zmet = loadPopColors.doall(indir, lib=lib)
id = inputDataDict["id"]
i = inputDataDict["i"]
ierr = inputDataDict["ierr"]
gr = inputDataDict["gr"]
ri = inputDataDict["ri"]
iz = inputDataDict["iz"]
grerr = inputDataDict["grerr"]
rierr = inputDataDict["rierr"]
izerr = inputDataDict["izerr"]
allzed = inputDataDict["zed"]
# protect against too small of errors => values = 0
ix = np.nonzero(grerr < 0.02)
grerr[ix] = 0.02
ix = np.nonzero(rierr < 0.02)
rierr[ix] = 0.02
ix = np.nonzero(izerr < 0.02)
izerr[ix] = 0.02
# prepping for output
out_id, out_zed, out_gr, out_stdgr, out_gi, out_stdgi, \
out_kri, out_stdkri, out_kii, out_stdkii, out_iobs, out_distmod, \
out_bestzmet,out_zmet,out_stdzmet, out_rabs, out_iabs, out_mass_gr, out_mass_gi, out_mass, out_stdmass, out_sfr, out_stdsfr, out_age, out_stdage = \
[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[],[]
out_bestsp = []
print "Computing stellar masses..."
size = id.size
for galaxy in range(0, size):
zed = allzed[galaxy]
print galaxy, " of ", size, " z = ", zed
rest_gr, rest_gi, weight = [], [], []
masslight, sfrs, ages, zmets, kii, kri = [], [], [], [], [], []
minChiSq = 999
spIndex = -1
for sp in range(0, len(splines)):
skey = str(sp) + "-" + str(zed)
sgr = splines[sp][0](zed)
sri = splines[sp][1](zed)
siz = splines[sp][2](zed)
sgrr = splines[sp][4](zed)
# restframe g-r
sgir = splines[sp][5](zed)
# restframe g-i
skii = splines[sp][6](zed)
# kcorrection: i_o - i_obs
skri = splines[sp][7](zed)
# kcorrection: r_o - i_obs
sml = splines[sp][8](zed)
# log(mass/light) (M_sun/L_sun)
ssfr = splines[sp][9](zed)
sage_cosmic = splines[sp][10](zed)
sage = splines[sp][11](zed)
szmet = zmet[sp]
#To be changed if SFH changes
gre = grerr[galaxy]
rie = rierr[galaxy]
ize = izerr[galaxy]
gr_chisq = pow((gr[galaxy] - sgr) / gre, 2)
ri_chisq = pow((ri[galaxy] - sri) / rie, 2)
iz_chisq = pow((iz[galaxy] - siz) / ize, 2)
rest_gr.append(sgrr)
rest_gi.append(sgir)
kii.append(skii)
kri.append(skri)
masslight.append(sml)
sfrs.append(ssfr)
ages.append(sage)
zmets.append(szmet)
chisq = gr_chisq + ri_chisq + iz_chisq
probability = 1 - chi2.cdf(chisq, 3 - 1)
# probability of chisq greater than this
weight.append(probability)
spIndex = np.argmax(weight)
rest_gr = np.array(rest_gr)
rest_gi = np.array(rest_gi)
kii = np.array(kii)
kri = np.array(kri)
masslight = np.array(masslight)
sfrs = np.array(sfrs)
ages = np.array(ages)
weight = np.array(weight)
gr_weighted = rest_gr * weight
gi_weighted = rest_gi * weight
kii_weighted = kii * weight
kri_weighted = kri * weight
masslight_weighted = masslight * weight
sfr_weighted = sfrs * weight
age_weighted = ages * weight
zmet_weighted = zmets * weight
w1 = weight.sum()
w2 = (weight**2).sum()
if w1 == 0: w1 = 1e-10
if w2 == 0: w2 = 1e-10
mean_gr = gr_weighted.sum() / w1
mean_gi = gi_weighted.sum() / w1
mean_kii = kii_weighted.sum() / w1
mean_kri = kri_weighted.sum() / w1
mean_masslight = masslight_weighted.sum() / w1
mean_sfr = sfr_weighted.sum() / w1
mean_age = age_weighted.sum() / w1
mean_zmet = zmet_weighted.sum() / w1
# unbiased weighted estimator of the sample variance
w3 = w1**2 - w2
if w3 == 0: w3 = 1e-10
var_gr = (w1 / w3) * (weight * (rest_gr - mean_gr)**2).sum()
var_gi = (w1 / w3) * (weight * (rest_gi - mean_gi)**2).sum()
var_kii = (w1 / w3) * (weight * (kii - mean_kii)**2).sum()
var_kri = (w1 / w3) * (weight * (kri - mean_kii)**2).sum()
var_masslight = (w1 / w3) * (weight *
(masslight - mean_masslight)**2).sum()
var_sfr = (w1 / w3) * (weight * (sfrs - mean_sfr)**2).sum()
var_age = (w1 / w3) * (weight * (ages - mean_age)**2).sum()
var_zmet = (w1 / w3) * (weight * (zmets - mean_zmet)**2).sum()
std_gr = var_gr**0.5
std_gi = var_gi**0.5
std_kii = var_kii**0.5
std_kri = var_kri**0.5
std_masslight = var_masslight**0.5
std_sfr = var_sfr**0.5
std_age = var_age**0.5
std_zmet = var_zmet**0.5
if std_gr > 99.99: std_gr = 99.99
if std_gi > 99.99: std_gi = 99.99
if std_kii > 99.99: std_kii = 99.99
if std_kri > 99.99: std_kri = 99.99
if std_sfr > 99.99: std_sfr = 99.99
if std_age > 99.99: std_age = 99.99
if std_masslight > 99.99: std_masslight = 99.99
if std_zmet > 99.99: std_zmet = 99.99
# Comment -distanceModulus out for fsps versions <2.5, as their mags don't include distance modulus
if zed in ldistDict:
lumdist = ldistDict[zed]
else:
lumdist = cd.luminosity_distance(zed)
# in Mpc
ldistDict[zed] = lumdist
distanceModulus = 5 * np.log10(lumdist * 1e6 / 10.)
iabs = i[galaxy] + mean_kii - distanceModulus
rabs = i[galaxy] + mean_kri - distanceModulus
taMass = taylorMass(mean_gi, iabs)
mcMass = mcintoshMass(mean_gr, rabs)
fsMass = fspsMass(mean_masslight, iabs)
# JTA: to make purely distance modulus
#iabs = i[galaxy] - distanceModulus
#fsMass = gstarMass( iabs )
# saving for output
out_id.append(id[galaxy])
out_gr.append(mean_gr)
out_stdgr.append(std_gr)
out_gi.append(mean_gi)
out_stdgi.append(std_gi)
out_kii.append(mean_kii)
out_stdkii.append(std_kii)
out_kri.append(mean_kri)
out_stdkri.append(std_kri)
out_iobs.append(i[galaxy])
out_distmod.append(distanceModulus)
out_iabs.append(iabs)
out_rabs.append(rabs)
out_mass_gr.append(mcMass)
out_mass_gi.append(taMass)
out_mass.append(fsMass)
out_stdmass.append(std_masslight)
out_bestsp.append(spIndex)
out_bestzmet.append(zmets[spIndex])
out_sfr.append(mean_sfr)
out_stdsfr.append(std_sfr)
out_age.append(mean_age)
out_stdage.append(std_age)
out_zmet.append(mean_zmet)
out_stdzmet.append(std_zmet)
out_zed.append(allzed[galaxy])
out_id = np.array(out_id).astype(int)
out_gr = np.array(out_gr)
out_stdgr = np.array(out_stdgr)
out_gi = np.array(out_gi)
out_stdgi = np.array(out_stdgi)
out_kii = np.array(out_kii)
out_stdkii = np.array(out_stdkii)
out_kri = np.array(out_kri)
out_stdkri = np.array(out_stdkri)
out_iobs = np.array(out_iobs)
out_distmod = np.array(out_distmod)
out_iabs = np.array(out_iabs)
out_rabs = np.array(out_rabs)
out_mass_gr = np.array(out_mass_gr)
out_mass_gi = np.array(out_mass_gi)
out_mass = np.array(out_mass)
out_stdmass = np.array(out_stdmass)
out_sfr = np.array(out_sfr)
out_stdsfr = np.array(out_stdsfr)
out_age = np.array(out_age)
out_stdage = np.array(out_stdage)
out_bestsp = np.array(out_bestsp)
out_zmet = np.array(out_zmet)
out_bestzmet = np.array(out_bestzmet)
out_zed = np.array(out_zed)
col1 = pf.Column(name='ID', format='K', array=out_id)
col2 = pf.Column(name='Z', format='E', array=out_zed)
col3 = pf.Column(name='gr_o', format='E', array=out_gr)
col4 = pf.Column(name='gr_o_err', format='E', array=out_stdgr)
col5 = pf.Column(name='gi_o', format='E', array=out_gi)
col6 = pf.Column(name='gi_o_err', format='E', array=out_stdgi)
col7 = pf.Column(name='kri', format='E', array=out_kri)
col8 = pf.Column(name='kri_err', format='E', array=out_stdkri)
col9 = pf.Column(name='kii', format='E', array=out_kii)
col10 = pf.Column(name='kii_err', format='E', array=out_stdkii)
col11 = pf.Column(name='distmod', format='E', array=out_distmod)
col12 = pf.Column(name='rabs', format='E', array=out_rabs)
col13 = pf.Column(name='iabs', format='E', array=out_iabs)
col14 = pf.Column(name='mass', format='E', array=out_mass)
col15 = pf.Column(name='mass_err', format='E', array=out_stdmass)
col16 = pf.Column(name='best_zmet', format='E', array=out_bestzmet)
col17 = pf.Column(name='zmet', format='E', array=out_zmet)
#col18=pf.Column(name='best_chisq',format='E',array=out_bestchisq)
#col20=pf.Column(name='ssfr',format='E',array=out_sfr)
#col21=pf.Column(name='ssfr_std',format='E',array=out_stdsfr)
#col22=pf.Column(name='mass_weight_age',format='E',array=out_age)
#col23=pf.Column(name='mass_weight_age_err',format='E',array=out_stdage)
#col24=pf.Column(name='best_model',format='E',array=out_bestsp)
cols = pf.ColDefs([
col1, col2, col3, col4, col5, col6, col7, col8, col9, col10, col11,
col12, col13, col14, col15, col16, col17
])
tbhdu = pf.BinTableHDU.from_columns(cols)
tbhdu.writeto(outfile + '.fits', clobber=True)
print "Output written to: ", outfile + "fits"