def compareGalpropzToBolshoiTrees(analyticalData,
BolshoiTrees,
redshifts,
h,
outputdir,
nvols=18,
no_phantoms=True,
galid=True):
#data storage
data = {}
#figure definitions
left, width = 0.1, 0.8
rect1 = [left, 0.1, width, 0.2]
rect2 = [left, 0.3, width, 0.65]
#note that this is only available on tuonela.stsci.edu
simuPath = '/Users/niemi/Desktop/Research/run/newtree1/'
simuDB = 'sams.db'
#star the figure
fig = P.figure()
ax1 = fig.add_axes(rect2) #left, bottom, width, height
ax2 = fig.add_axes(rect1)
#set title
if galid:
ax1.set_title(
'Dark Matter Halo Mass Functions (gal\_id = 1; 26 volumes)')
else:
ax1.set_title('Dark Matter Halo Mass Functions (galpropz: 26 volumes)')
#loop over the data and redshift range
for redsh, BolshoiTree, anaData in zip(sorted(redshifts.itervalues()),
BolshoiTrees, analyticalData):
#skip some redshifts
if redsh < 1.0 or redsh == 2.0276 or redsh == 5.1614\
or redsh == 6.5586 or redsh == 3.0584:
continue
#if redsh == 1.0064 or redsh == 3.0584 or redsh == 4.0429 or \
# redsh == 8.2251:
# continue
#change this to logging afterwords
logging.debug(redsh)
logging.debug(BolshoiTree)
logging.debug(anaData)
rlow = redsh - 0.02
rhigh = redsh + 0.02
if galid:
query = '''select mhalo from galpropz where
galpropz.zgal > {0:f} and galpropz.zgal < {1:f} and
galpropz.gal_id = 1'''.format(rlow, rhigh)
else:
query = '''select mhalo from galpropz where
galpropz.zgal > {0:f} and
galpropz.zgal < {1:f}'''.format(rlow, rhigh)
logging.debug(query)
data['SAM'] = db.sqlite.get_data_sqlite(simuPath, simuDB, query) * 1e9
#calculate the mass functions from the SAM data, only x volumes
mbin0SAM, mf0SAM = df.diffFunctionLogBinning(data['SAM'],
nbins=30,
h=h,
mmin=1e9,
mmax=1e15,
volume=50.0,
nvols=nvols,
physical_units=True)
mbin0SAM = 10**mbin0SAM
#mf0SAM = mf0SAM * 1.7
#read the Bolshoi merger trees
data['Bolshoi'] = io.readBolshoiDMfile(BolshoiTree, 0, no_phantoms)
#calculate the mass functions from the full Bolshoi data
mbin0Bolshoi, mf0Bolshoi = df.diffFunctionLogBinning(
data['Bolshoi'] / h,
nbins=30,
h=h,
mmin=1e9,
mmax=1e15,
volume=50.0,
nvols=125,
physical_units=True)
mbin0Bolshoi = 10**mbin0Bolshoi
#Analytical MFs
#get Rachel's analytical curves
#M dN/dM dNcorr/dM dN/dlog10(M) dN/dlog10(Mcorr)
d = N.loadtxt(anaData)
data['Press-Schecter'] = N.array([d[:, 0], d[:, 3]])
data['Sheth-Tormen'] = N.array([d[:, 0], d[:, 4]])
#ST
sh = ax1.plot(data['Sheth-Tormen'][0],
data['Sheth-Tormen'][1],
'k-',
lw=0.9)
#PS
#ps = ax1.plot(data['Press-Schecter'][0],
# data['Press-Schecter'][1],
# 'g--', lw = 1.1)
#MF from Bolshoi
bolshoiax = ax1.plot(mbin0Bolshoi, mf0Bolshoi, 'ro--', ms=4)
#MF from the SAM run
samax = ax1.plot(mbin0SAM, mf0SAM, 'gs--', ms=4)
#mark redshift
for a, b in zip(mbin0Bolshoi[::-1], mf0Bolshoi[::-1]):
if b > 10**-5:
break
ax1.annotate('$z \sim {0:.2f}$'.format(redsh), (0.6 * a, 3 * 10**-6),
size='x-small')
#plot the residuals
if redsh < 1.5:
#make the plot
ax2.annotate('$z \sim {0:.2f}$'.format(redsh), (1.5 * 10**9, 1.05),
xycoords='data',
size=10)
ax2.axhline(1.0, color='k')
msk = mf0SAM / mf0Bolshoi > 0
ax2.plot(mbin0SAM[msk], mf0SAM[msk] / mf0Bolshoi[msk], 'r-')
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_ylim(1e-6, 10**-0)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(2e9, 4e14)
ax2.set_xlim(2e9, 4e14)
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(
r'$\mathrm{d}N / \mathrm{d}\log_{10}(M_{\mathrm{vir}}) \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$'
)
ax2.set_ylabel(r'$\frac{\mathrm{galpropz.dat}}{\mathrm{IsoTree}}$')
ax1.legend((sh, bolshoiax, samax), ('Sheth-Tormen', 'Bolshoi', 'galpropz'),
shadow=True,
fancybox=True,
numpoints=1)
if galid:
P.savefig(outputdir + 'IsotreesVSgalpropzGalID.pdf')
else:
P.savefig(outputdir + 'IsotreesVSgalpropz.pdf')
def plot_mass_function(redshift, h, no_phantoms, *data):
#fudge factor
fudge = 2.0
#http://adsabs.harvard.edu/abs/2001MNRAS.321..372J
#Jenkins et al. paper has sqrt(2./pi) while
#Rachel's code has 1/(sqrt(2*pi))
#ratio of these two is the fudge factor
#read data to dictionary
dt = {}
for x in data[0]:
if 'Bolshoi' in x:
dt[x] = io.readBolshoiDMfile(data[0][x], 0, no_phantoms)
else:
dt[x] = N.loadtxt(data[0][x])
#calculate the mass functions from the Bolshoi data
mbin0, mf0 = df.diffFunctionLogBinning(dt['Bolshoi'] / h,
nbins=35,
h=0.7,
mmin=10**9.0,
mmax=10**15.0,
physical_units=True)
del dt['Bolshoi']
#use chain rule to get dN / dM
#dN/dM = dN/dlog10(M) * dlog10(M)/dM
#d/dM (log10(M)) = 1 / (M*ln(10))
mf0 *= 1. / (mbin0 * N.log(10))
#put mass back to power
mbin0 = 10**mbin0
#title
if no_phantoms:
ax1.set_title('Bolshoi Dark Matter Mass Functions (no phantoms)')
else:
ax1.set_title('Bolshoi Dark Matter Mass Functions')
#mark redshift
for a, b in zip(mbin0[::-1], mf0[::-1]):
if b > 10**-6:
break
ax1.annotate('$z \sim %.1f$' % redshift, (0.98 * a, 3 * 10**-6),
size='x-small')
#Analytical MFs
#0th column: log10 of mass (Msolar, NOT Msolar/h)
#1st column: mass (Msolar/h)
#2nd column: (dn/dM)*dM, per Mpc^3 (NOT h^3/Mpc^3)
xST = 10**dt['Sheth-Tormen'][:, 0]
yST = dt['Sheth-Tormen'][:, 2] * fudge
sh = ax1.plot(xST, yST, 'b-', lw=1.3)
#PS
xPS = 10**dt['Press-Schecter'][:, 0]
yPS = dt['Press-Schecter'][:, 2] * fudge
ps = ax1.plot(xPS, yPS, 'g--', lw=1.1)
#MF from Bolshoi
bolshoi = ax1.plot(mbin0, mf0, 'ro:', ms=5)
#delete data to save memory, dt is not needed any longer
del dt
#plot the residuals
if round(float(redshift), 1) < 1.5:
#interploate to right x scale
yST = N.interp(mbin0, xST, yST)
yPS = N.interp(mbin0, xPS, yPS)
#make the plot
ax2.annotate('$z \sim %.1f$' % redshift, (1.5 * 10**9, 1.05),
xycoords='data',
size=10)
ax2.axhline(1.0, color='b')
ax2.plot(mbin0, mf0 / yST, 'b-')
ax2.plot(mbin0, mf0 / yPS, 'g-')
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_ylim(10**-7, 10**-1)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(10**9, 10**15)
ax2.set_xlim(10**9, 10**15)
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(
r'$\mathrm{d}N / \mathrm{d}M_{\mathrm{vir}} \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$'
)
ax2.set_ylabel(r'$\frac{\mathrm{Bolshoi}}{\mathrm{Model}}$')
ax1.legend((bolshoi, sh, ps),
('Bolshoi', 'Sheth-Tormen', 'Press-Schecter'),
shadow=True,
fancybox=True,
numpoints=1)
def plotDMMFfromGalpropzAnalytical2(redshift, h, *data):
#fudge factor
fudge = 1.
#http://adsabs.harvard.edu/abs/2001MNRAS.321..372J
#Jenkins et al. paper has sqrt(2./pi) while
#Rachel's code has 1/(sqrt(2*pi))
#ratio of these two is the fudge factor
#find the home directory, because the output is to dropbox
#and my user name is not always the same, this hack is required.
hm = os.getenv('HOME')
path = hm + '/Dropbox/Research/Bolshoi/run/trial2/'
database = 'sams.db'
rlow = redshift - 0.1
rhigh = redshift + 0.1
query = '''select mhalo from galpropz where
galpropz.zgal > %f and galpropz.zgal <= %f and
galpropz.gal_id = 1
''' % (rlow, rhigh)
#Hubble constants
h3 = h**3
#read data
dt = {}
for x in data[0]:
#M dN/dM dNcorr/dM dN/dlog10(M) dN/dlog10(Mcorr)
d = N.loadtxt(data[0][x])
dt['Press-Schecter'] = N.array([d[:, 0], d[:, 3]])
dt['Sheth-Tormen'] = N.array([d[:, 0], d[:, 4]])
dt['Bolshoi'] = db.sqlite.get_data_sqlite(path, database, query) * 1e9
print len(dt['Bolshoi'])
#calculate the mass functions from the Bolshoi data
mbin0, mf0 = df.diffFunctionLogBinning(dt['Bolshoi'] / h,
nbins=35,
h=0.7,
mmin=10**9.0,
mmax=10**15.0,
volume=50,
nvols=26,
physical_units=True)
del dt['Bolshoi']
mbin0 = 10**mbin0
#title
ax1.set_title('Dark Matter Halo Mass Functions (galpropz.dat)')
#mark redshift
for a, b in zip(mbin0[::-1], mf0[::-1]):
if b > 10**-5:
break
ax1.annotate('$z \sim %.0f$' % redshift, (0.98 * a, 3 * 10**-6),
size='x-small')
#Analytical MFs
xST = dt['Sheth-Tormen'][0]
yST = dt['Sheth-Tormen'][1] * fudge
sh = ax1.plot(xST, yST, 'b-', lw=1.3)
#PS
xPS = dt['Press-Schecter'][0]
yPS = dt['Press-Schecter'][1] * fudge
ps = ax1.plot(xPS, yPS, 'g--', lw=1.1)
#MF from Bolshoi
bolshoi = ax1.plot(mbin0, mf0, 'ro:', ms=5)
#delete data to save memory, dt is not needed any longer
del dt
#plot the residuals
if redshift < 1.5:
#interploate to right x scale
mfintST = N.interp(xST, mbin0, mf0)
mfintPS = N.interp(xPS, mbin0, mf0)
#yST = N.interp(mbin0, xST, yST)
#yPS = N.interp(mbin0, xPS, yPS)
#make the plot
ax2.annotate('$z \sim %.0f$' % redshift, (1.5 * 10**9, 1.05),
xycoords='data',
size=10)
ax2.axhline(1.0, color='b')
ax2.plot(xST, mfintST / yST, 'b-')
ax2.plot(xPS, mfintPS / yPS, 'g-')
#ax2.plot(mbin0, mf0 / yST, 'b-')
#ax2.plot(mbin0, mf0 / yPS, 'g-')
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_ylim(10**-6, 10**-0)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(10**9, 10**15)
ax2.set_xlim(10**9, 10**15)
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(
r'$\mathrm{d}N / \mathrm{d}\log_{10}(M_{\mathrm{vir}}) \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$'
)
ax2.set_ylabel(r'$\frac{\mathrm{galpropz.dat}}{\mathrm{Model}}$')
ax1.legend((bolshoi, sh, ps),
('Bolshoi', 'Sheth-Tormen', 'Press-Schecter'),
shadow=True,
fancybox=True,
numpoints=1)
def plot_mass_functionAnalytical2(redshift, h, no_phantoms, *data):
#fudge factor
fudge = 1.
#http://adsabs.harvard.edu/abs/2001MNRAS.321..372J
#Jenkins et al. paper has sqrt(2./pi) while
#Rachel's code has 1/(sqrt(2*pi))
#ratio of these two is the fudge factor
#read data
dt = {}
for x in data[0]:
if 'Bolshoi' in x:
dt[x] = io.readBolshoiDMfile(data[0][x], 0, no_phantoms)
else:
#M dN/dM dNcorr/dM dN/dlog10(M) dN/dlog10(Mcorr)
d = N.loadtxt(data[0][x])
dt['Press-Schecter'] = N.array([d[:, 0], d[:, 3]])
dt['Sheth-Tormen'] = N.array([d[:, 0], d[:, 4]])
#calculate the mass functions from the Bolshoi data
mbin0, mf0 = df.diffFunctionLogBinning(dt['Bolshoi'] / h,
nbins=35,
h=0.7,
mmin=10**9.0,
mmax=10**15.0,
physical_units=True)
del dt['Bolshoi']
mbin0 = 10**mbin0
#title
if no_phantoms:
ax1.set_title('Bolshoi Dark Matter Mass Functions (no phantoms)')
else:
ax1.set_title('Bolshoi Dark Matter Mass Functions')
#mark redshift
for a, b in zip(mbin0[::-1], mf0[::-1]):
if b > 10**-6:
break
ax1.annotate('$z \sim %.1f$' % redshift, (0.98 * a, 3 * 10**-6),
size='x-small')
#Analytical MFs
xST = dt['Sheth-Tormen'][0]
yST = dt['Sheth-Tormen'][1] * fudge
print xST[1000], yST[1000]
sh = ax1.plot(xST, yST, 'b-', lw=1.3)
#PS
xPS = dt['Press-Schecter'][0]
yPS = dt['Press-Schecter'][1] * fudge
ps = ax1.plot(xPS, yPS, 'g--', lw=1.1)
#MF from Bolshoi
bolshoi = ax1.plot(mbin0, mf0, 'ro:', ms=5)
#delete data to save memory, dt is not needed any longer
del dt
#plot the residuals
if round(float(redshift), 1) < 1.5:
#interploate to right x scale
mfintST = N.interp(xST, mbin0, mf0)
mfintPS = N.interp(xPS, mbin0, mf0)
#yST = N.interp(mbin0, xST, yST)
#yPS = N.interp(mbin0, xPS, yPS)
#make the plot
ax2.annotate('$z \sim %.0f$' % redshift, (1.5 * 10**9, 1.05),
xycoords='data',
size=10)
ax2.axhline(1.0, color='b')
ax2.plot(xST, mfintST / yST, 'b-')
ax2.plot(xPS, mfintPS / yPS, 'g-')
#ax2.plot(mbin0, mf0 / yST, 'b-')
#ax2.plot(mbin0, mf0 / yPS, 'g-')
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_ylim(3 * 10**-7, 10**0)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(10**9, 10**15)
ax2.set_xlim(10**9, 10**15)
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(
r'$\mathrm{d}N / \mathrm{d}\log_{10}(M_{\mathrm{vir}}) \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$'
)
ax2.set_ylabel(r'$\frac{\mathrm{Bolshoi}}{\mathrm{Model}}$')
ax1.legend((bolshoi, sh, ps),
('Bolshoi', 'Sheth-Tormen', 'Press-Schecter'),
shadow=True,
fancybox=True,
numpoints=1)
def plotDMMFfromGalpropz(redshift, h, *data):
#fudge factor
fudge = 2.0
#http://adsabs.harvard.edu/abs/2001MNRAS.321..372J
#Jenkins et al. paper has sqrt(2./pi) while
#Rachel's code has 1/(sqrt(2*pi))
#ratio of these two is the fudge factor
#find the home directory, because the output is to dropbox
#and my user name is not always the same, this hack is required.
hm = os.getenv('HOME')
path = hm + '/Dropbox/Research/Bolshoi/run/trial2/'
database = 'sams.db'
rlow = redshift - 0.1
rhigh = redshift + 0.1
query = '''select mhalo from galpropz where
galpropz.zgal > %f and galpropz.zgal <= %f and
galpropz.gal_id = 1
''' % (rlow, rhigh)
print query
#read data
dt = {}
for x in data[0]:
dt[x] = N.loadtxt(data[0][x])
dt['Bolshoi'] = db.sqlite.get_data_sqlite(path, database, query) * 1e9
#calculate the mass functions from the Bolshoi data
mbin0, mf0 = df.diffFunctionLogBinning(dt['Bolshoi'] / h,
nbins=35,
h=0.7,
mmin=10**9.0,
mmax=10**15.0,
volume=50,
nvols=26,
physical_units=True)
del dt['Bolshoi']
#use chain rule to get dN / dM
#dN/dM = dN/dlog10(M) * dlog10(M)/dM
#d/dM (log10(M)) = 1 / (M*ln(10))
mf0 *= 1. / (mbin0 * N.log(10))
mbin0 = 10**mbin0
#title
ax1.set_title('Dark Matter Halo Mass Functions (galpropz.dat)')
#mark redshift
for a, b in zip(mbin0[::-1], mf0[::-1]):
if b > 10**-6:
break
ax1.annotate('$z \sim %.0f$' % redshift, (0.98 * a, 3 * 10**-7),
size='x-small')
#Analytical MFs
#0th column: log10 of mass (Msolar, NOT Msolar/h)
#1st column: mass (Msolar/h)
#2nd column: (dn/dM)*dM, per Mpc^3 (NOT h^3/Mpc^3)
xST = 10**dt['Sheth-Tormen'][:, 0]
yST = dt['Sheth-Tormen'][:, 2] * fudge
sh = ax1.plot(xST, yST, 'b-', lw=1.3)
#PS
xPS = 10**dt['Press-Schecter'][:, 0]
yPS = dt['Press-Schecter'][:, 2] * fudge
ps = ax1.plot(xPS, yPS, 'g--', lw=1.1)
#MF from Bolshoi
bolshoi = ax1.plot(mbin0, mf0, 'ro:', ms=5)
#delete data to save memory, dt is not needed any longer
del dt
#plot the residuals
if round(float(redshift), 1) < 1.5:
#interploate to right x scale
ySTint = N.interp(mbin0, xST, yST)
yPSint = N.interp(mbin0, xPS, yPS)
#make the plot
ax2.annotate('$z \sim %.0f$' % redshift, (1.5 * 10**9, 1.05),
xycoords='data',
size=10)
ax2.axhline(1.0, color='b')
ax2.plot(mbin0, mf0 / ySTint, 'b-')
ax2.plot(mbin0, mf0 / yPSint, 'g-')
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_ylim(5 * 10**-8, 10**-1)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(10**9, 10**15)
ax2.set_xlim(10**9, 10**15)
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(
r'$\mathrm{d}N / \mathrm{d}M_{\mathrm{vir}} \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$'
)
ax2.set_ylabel(r'$\frac{\mathrm{galpropz.dat}}{\mathrm{Model}}$')
ax1.legend((bolshoi, sh, ps),
('Bolshoi', 'Sheth-Tormen', 'Press-Schecter'),
shadow=True,
fancybox=True,
numpoints=1)
def compareGalpropzToBolshoiTrees(analyticalData,
BolshoiTrees,
redshifts,
h,
outputdir,
nvols=18,
no_phantoms=True,
galid=True):
#data storage
data = {}
#figure definitions
left, width = 0.1, 0.8
rect1 = [left, 0.1, width, 0.2]
rect2 = [left, 0.3, width, 0.65]
#note that this is only available on tuonela.stsci.edu
simuPath = '/Users/niemi/Desktop/Research/run/newtree1/'
simuDB = 'sams.db'
#star the figure
fig = P.figure()
ax1 = fig.add_axes(rect2) #left, bottom, width, height
ax2 = fig.add_axes(rect1)
#set title
if galid:
ax1.set_title('Dark Matter Halo Mass Functions (gal\_id = 1; 26 volumes)')
else:
ax1.set_title('Dark Matter Halo Mass Functions (galpropz: 26 volumes)')
#loop over the data and redshift range
for redsh, BolshoiTree, anaData in zip(sorted(redshifts.itervalues()),
BolshoiTrees,
analyticalData):
#skip some redshifts
if redsh < 1.0 or redsh == 2.0276 or redsh == 5.1614\
or redsh == 6.5586 or redsh == 3.0584:
continue
#if redsh == 1.0064 or redsh == 3.0584 or redsh == 4.0429 or \
# redsh == 8.2251:
# continue
#change this to logging afterwords
logging.debug(redsh)
logging.debug(BolshoiTree)
logging.debug(anaData)
rlow = redsh - 0.02
rhigh = redsh + 0.02
if galid:
query = '''select mhalo from galpropz where
galpropz.zgal > {0:f} and galpropz.zgal < {1:f} and
galpropz.gal_id = 1'''.format(rlow, rhigh)
else:
query = '''select mhalo from galpropz where
galpropz.zgal > {0:f} and
galpropz.zgal < {1:f}'''.format(rlow, rhigh)
logging.debug(query)
data['SAM'] = db.sqlite.get_data_sqlite(simuPath, simuDB, query) * 1e9
#calculate the mass functions from the SAM data, only x volumes
mbin0SAM, mf0SAM = df.diffFunctionLogBinning(data['SAM'],
nbins=30,
h=h,
mmin=1e9,
mmax=1e15,
volume=50.0,
nvols=nvols,
physical_units=True)
mbin0SAM = 10 ** mbin0SAM
#mf0SAM = mf0SAM * 1.7
#read the Bolshoi merger trees
data['Bolshoi'] = io.readBolshoiDMfile(BolshoiTree, 0, no_phantoms)
#calculate the mass functions from the full Bolshoi data
mbin0Bolshoi, mf0Bolshoi = df.diffFunctionLogBinning(data['Bolshoi'] / h,
nbins=30,
h=h,
mmin=1e9,
mmax=1e15,
volume=50.0,
nvols=125,
physical_units=True)
mbin0Bolshoi = 10 ** mbin0Bolshoi
#Analytical MFs
#get Rachel's analytical curves
#M dN/dM dNcorr/dM dN/dlog10(M) dN/dlog10(Mcorr)
d = N.loadtxt(anaData)
data['Press-Schecter'] = N.array([d[:, 0], d[:, 3]])
data['Sheth-Tormen'] = N.array([d[:, 0], d[:, 4]])
#ST
sh = ax1.plot(data['Sheth-Tormen'][0],
data['Sheth-Tormen'][1],
'k-', lw=0.9)
#PS
#ps = ax1.plot(data['Press-Schecter'][0],
# data['Press-Schecter'][1],
# 'g--', lw = 1.1)
#MF from Bolshoi
bolshoiax = ax1.plot(mbin0Bolshoi,
mf0Bolshoi,
'ro--', ms=4)
#MF from the SAM run
samax = ax1.plot(mbin0SAM,
mf0SAM,
'gs--', ms=4)
#mark redshift
for a, b in zip(mbin0Bolshoi[::-1], mf0Bolshoi[::-1]):
if b > 10 ** -5:
break
ax1.annotate('$z \sim {0:.2f}$'.format(redsh),
(0.6 * a, 3 * 10 ** -6), size='x-small')
#plot the residuals
if redsh < 1.5:
#make the plot
ax2.annotate('$z \sim {0:.2f}$'.format(redsh),
(1.5 * 10 ** 9, 1.05), xycoords='data',
size=10)
ax2.axhline(1.0, color='k')
msk = mf0SAM / mf0Bolshoi > 0
ax2.plot(mbin0SAM[msk],
mf0SAM[msk] / mf0Bolshoi[msk],
'r-')
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_ylim(1e-6, 10 ** -0)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(2e9, 4e14)
ax2.set_xlim(2e9, 4e14)
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(r'$\mathrm{d}N / \mathrm{d}\log_{10}(M_{\mathrm{vir}}) \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$')
ax2.set_ylabel(r'$\frac{\mathrm{galpropz.dat}}{\mathrm{IsoTree}}$')
ax1.legend((sh, bolshoiax, samax),
('Sheth-Tormen', 'Bolshoi', 'galpropz'),
shadow=True, fancybox=True,
numpoints=1)
if galid:
P.savefig(outputdir + 'IsotreesVSgalpropzGalID.pdf')
else:
P.savefig(outputdir + 'IsotreesVSgalpropz.pdf')
def plotMassFunction(simus,
sheth,
outdir,
h,
simuPath,
simuDB,
noPhantoms=True):
'''
This function plots dark matter halo mass functions from
file(s) given in simus. The usual usage is that simus
contain a list of file names. These files should contain
dark matter halo masses derived from the Bolshoi isotree
files.
This function uses a fudge factor = 2.0 because
Jenkins et al. paper
(http://adsabs.harvard.edu/abs/2001MNRAS.321..372J)
has sqrt(2./pi) while
Rachel's code has 1/(sqrt(2*pi))
The ratio of these two definitions is the fudge factor.
'''
#fudge factor
fudge = 2.0
#figure definitions
left, width = 0.1, 0.8
rect1 = [left, 0.1, width, 0.2]
rect2 = [left, 0.3, width, 0.65]
for i, file in enumerate(simus):
#get the redshift
redshift = float(file.split('z')[1][:5])
logging.debug('Plotting redshift %.2f dark matter mass functions',
redshift)
#different redshift ranges
if redshift < 1.8:
rlow = redshift - 0.02
rhigh = redshift + 0.02
else:
rlow = redshift - 0.1
rhigh = redshift + 0.1
#read bolshoi data
BoshoiData = io.readBolshoiDMfile(file, 0, noPhantoms)
#calculate the mass functions from the Bolshoi data
mbin0, mf0 = df.diffFunctionLogBinning(BoshoiData / h,
nbins=35,
h=h,
mmin=1e9,
mmax=1e15,
volume=50.0,
nvols=18)
#use chain rule to get dN / dM
#dN/dM = dN/dlog10(M) * dlog10(M)/dM
#d/dM (log10(M)) = 1 / (M*ln(10))
mf0 *= 1. / (mbin0 * N.log(10))
#put mass back to power
mbin0 = 10**mbin0
#get the SAMs data
#if round(redshift, 2) < 0.08:
if i < 1:
query = '''select mhalo from galprop where galprop.gal_id = 1'''
GalpropData = 10**db.sqlite.get_data_sqlite(
simuPath, simuDB, query)
else:
query = '''select mhalo from galpropz where
galpropz.zgal > {0:f} and galpropz.zgal < {1:f} and
galpropz.gal_id = 1'''.format(rlow, rhigh)
GalpropData = db.sqlite.get_data_sqlite(simuPath, simuDB,
query) * 1e9
logging.debug(query)
#calculate the mass functions from the SAM data, only x volumes
mbin0SAM, mf0SAM = df.diffFunctionLogBinning(GalpropData,
nbins=35,
h=h,
mmin=1e9,
mmax=1e15,
volume=50.0,
nvols=15)
#use chain rule to get dN / dM
#dN/dM = dN/dlog10(M) * dlog10(M)/dM
#d/dM (log10(M)) = 1 / (M*ln(10))
mf0SAM *= 1. / (mbin0SAM * N.log(10))
mbin0SAM = 10**mbin0SAM
#start a figure
fig = P.figure(figsize=(10, 10))
ax1 = fig.add_axes(rect2) #left, bottom, width, height
ax2 = fig.add_axes(rect1)
#title
if noPhantoms:
ax1.set_title('$z \sim %.2f$ (no phantoms)' % redshift)
else:
ax1.set_title('$z \sim %.2f$' % redshift)
#Analytical MFs
#0th column: log10 of mass (Msolar, NOT Msolar/h)
#1st column: mass (Msolar/h)
#2nd column: (dn/dM)*dM, per Mpc^3 (NOT h^3/Mpc^3)
logging.debug(sheth[i])
dt = N.loadtxt(sheth[i])
xST = 10**dt[:, 0]
yST = dt[:, 2] * fudge
sh = ax1.plot(xST, yST, 'b-', lw=1.3)
#MF from Bolshoi
bolshoi = ax1.plot(mbin0, mf0, 'ro:', ms=5)
#MF from the SAM run
samax = ax1.plot(mbin0SAM, mf0SAM, 'gs--', ms=4)
#plot the residuals
ax2.axhline(1.0, color='k')
ax2.plot(mbin0, mf0 / mf0SAM, 'm-')
#interploate to right x scale
mfintST1 = N.interp(xST, mbin0, mf0)
mfintST2 = N.interp(xST, mbin0SAM, mf0SAM)
ax2.plot(xST, mfintST1 / yST, 'r-')
ax2.plot(xST, mfintST2 / yST, 'g-')
ax1.set_ylim(1e-7, 1e-1)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(1e9, 1e15)
ax2.set_xlim(1e9, 1e15)
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(
r'$\mathrm{d}N / \mathrm{d}M_{\mathrm{vir}} \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$'
)
ax2.set_ylabel(r'$\frac{\mathrm{Bolshoi}}{\mathrm{Model}}$')
ax1.legend((bolshoi, samax, sh),
('Bolshoi', 'galpropz', 'Sheth-Tormen'),
shadow=True,
fancybox=True,
numpoints=1)
P.savefig(outdir + 'DMMF{0:d}.png'.format(i))
P.close()
def plotDMMFfromGalpropzAnalytical2(redshift, h, *data):
#fudge factor
fudge = 1.
#http://adsabs.harvard.edu/abs/2001MNRAS.321..372J
#Jenkins et al. paper has sqrt(2./pi) while
#Rachel's code has 1/(sqrt(2*pi))
#ratio of these two is the fudge factor
#find the home directory, because the output is to dropbox
#and my user name is not always the same, this hack is required.
hm = os.getenv('HOME')
path = hm + '/Dropbox/Research/Bolshoi/run/trial2/'
database = 'sams.db'
rlow = redshift - 0.1
rhigh = redshift + 0.1
query = '''select mhalo from galpropz where
galpropz.zgal > %f and galpropz.zgal <= %f and
galpropz.gal_id = 1
''' % (rlow, rhigh)
#Hubble constants
h3 = h ** 3
#read data
dt = {}
for x in data[0]:
#M dN/dM dNcorr/dM dN/dlog10(M) dN/dlog10(Mcorr)
d = N.loadtxt(data[0][x])
dt['Press-Schecter'] = N.array([d[:, 0], d[:, 3]])
dt['Sheth-Tormen'] = N.array([d[:, 0], d[:, 4]])
dt['Bolshoi'] = db.sqlite.get_data_sqlite(path, database, query) * 1e9
print len(dt['Bolshoi'])
#calculate the mass functions from the Bolshoi data
mbin0, mf0 = df.diffFunctionLogBinning(dt['Bolshoi'] / h,
nbins=35,
h=0.7,
mmin=10 ** 9.0,
mmax=10 ** 15.0,
volume=50,
nvols=26,
physical_units=True)
del dt['Bolshoi']
mbin0 = 10 ** mbin0
#title
ax1.set_title('Dark Matter Halo Mass Functions (galpropz.dat)')
#mark redshift
for a, b in zip(mbin0[::-1], mf0[::-1]):
if b > 10 ** -5:
break
ax1.annotate('$z \sim %.0f$' % redshift,
(0.98 * a, 3 * 10 ** -6), size='x-small')
#Analytical MFs
xST = dt['Sheth-Tormen'][0]
yST = dt['Sheth-Tormen'][1] * fudge
sh = ax1.plot(xST, yST, 'b-', lw=1.3)
#PS
xPS = dt['Press-Schecter'][0]
yPS = dt['Press-Schecter'][1] * fudge
ps = ax1.plot(xPS, yPS, 'g--', lw=1.1)
#MF from Bolshoi
bolshoi = ax1.plot(mbin0, mf0, 'ro:', ms=5)
#delete data to save memory, dt is not needed any longer
del dt
#plot the residuals
if redshift < 1.5:
#interploate to right x scale
mfintST = N.interp(xST, mbin0, mf0)
mfintPS = N.interp(xPS, mbin0, mf0)
#yST = N.interp(mbin0, xST, yST)
#yPS = N.interp(mbin0, xPS, yPS)
#make the plot
ax2.annotate('$z \sim %.0f$' % redshift,
(1.5 * 10 ** 9, 1.05), xycoords='data',
size=10)
ax2.axhline(1.0, color='b')
ax2.plot(xST, mfintST / yST, 'b-')
ax2.plot(xPS, mfintPS / yPS, 'g-')
#ax2.plot(mbin0, mf0 / yST, 'b-')
#ax2.plot(mbin0, mf0 / yPS, 'g-')
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_ylim(10 ** -6, 10 ** -0)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(10 ** 9, 10 ** 15)
ax2.set_xlim(10 ** 9, 10 ** 15)
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(r'$\mathrm{d}N / \mathrm{d}\log_{10}(M_{\mathrm{vir}}) \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$')
ax2.set_ylabel(r'$\frac{\mathrm{galpropz.dat}}{\mathrm{Model}}$')
ax1.legend((bolshoi, sh, ps),
('Bolshoi', 'Sheth-Tormen', 'Press-Schecter'),
shadow=True, fancybox=True,
numpoints=1)
def plot_mass_function(redshift, h, no_phantoms, *data):
#fudge factor
fudge = 2.0
#http://adsabs.harvard.edu/abs/2001MNRAS.321..372J
#Jenkins et al. paper has sqrt(2./pi) while
#Rachel's code has 1/(sqrt(2*pi))
#ratio of these two is the fudge factor
#read data to dictionary
dt = {}
for x in data[0]:
if 'Bolshoi' in x:
dt[x] = io.readBolshoiDMfile(data[0][x], 0, no_phantoms)
else:
dt[x] = N.loadtxt(data[0][x])
#calculate the mass functions from the Bolshoi data
mbin0, mf0 = df.diffFunctionLogBinning(dt['Bolshoi'] / h,
nbins=35,
h=0.7,
mmin=10 ** 9.0,
mmax=10 ** 15.0,
physical_units=True)
del dt['Bolshoi']
#use chain rule to get dN / dM
#dN/dM = dN/dlog10(M) * dlog10(M)/dM
#d/dM (log10(M)) = 1 / (M*ln(10))
mf0 *= 1. / (mbin0 * N.log(10))
#put mass back to power
mbin0 = 10 ** mbin0
#title
if no_phantoms:
ax1.set_title('Bolshoi Dark Matter Mass Functions (no phantoms)')
else:
ax1.set_title('Bolshoi Dark Matter Mass Functions')
#mark redshift
for a, b in zip(mbin0[::-1], mf0[::-1]):
if b > 10 ** -6:
break
ax1.annotate('$z \sim %.1f$' % redshift,
(0.98 * a, 3 * 10 ** -6), size='x-small')
#Analytical MFs
#0th column: log10 of mass (Msolar, NOT Msolar/h)
#1st column: mass (Msolar/h)
#2nd column: (dn/dM)*dM, per Mpc^3 (NOT h^3/Mpc^3)
xST = 10 ** dt['Sheth-Tormen'][:, 0]
yST = dt['Sheth-Tormen'][:, 2] * fudge
sh = ax1.plot(xST, yST, 'b-', lw=1.3)
#PS
xPS = 10 ** dt['Press-Schecter'][:, 0]
yPS = dt['Press-Schecter'][:, 2] * fudge
ps = ax1.plot(xPS, yPS, 'g--', lw=1.1)
#MF from Bolshoi
bolshoi = ax1.plot(mbin0, mf0, 'ro:', ms=5)
#delete data to save memory, dt is not needed any longer
del dt
#plot the residuals
if round(float(redshift), 1) < 1.5:
#interploate to right x scale
yST = N.interp(mbin0, xST, yST)
yPS = N.interp(mbin0, xPS, yPS)
#make the plot
ax2.annotate('$z \sim %.1f$' % redshift,
(1.5 * 10 ** 9, 1.05), xycoords='data',
size=10)
ax2.axhline(1.0, color='b')
ax2.plot(mbin0, mf0 / yST, 'b-')
ax2.plot(mbin0, mf0 / yPS, 'g-')
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_ylim(10 ** -7, 10 ** -1)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(10 ** 9, 10 ** 15)
ax2.set_xlim(10 ** 9, 10 ** 15)
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(r'$\mathrm{d}N / \mathrm{d}M_{\mathrm{vir}} \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$')
ax2.set_ylabel(r'$\frac{\mathrm{Bolshoi}}{\mathrm{Model}}$')
ax1.legend((bolshoi, sh, ps),
('Bolshoi', 'Sheth-Tormen', 'Press-Schecter'),
shadow=True, fancybox=True,
numpoints=1)
def plotDMMFfromGalpropz(redshift, h, *data):
#fudge factor
fudge = 2.0
#http://adsabs.harvard.edu/abs/2001MNRAS.321..372J
#Jenkins et al. paper has sqrt(2./pi) while
#Rachel's code has 1/(sqrt(2*pi))
#ratio of these two is the fudge factor
#find the home directory, because the output is to dropbox
#and my user name is not always the same, this hack is required.
hm = os.getenv('HOME')
path = hm + '/Dropbox/Research/Bolshoi/run/trial2/'
database = 'sams.db'
rlow = redshift - 0.1
rhigh = redshift + 0.1
query = '''select mhalo from galpropz where
galpropz.zgal > %f and galpropz.zgal <= %f and
galpropz.gal_id = 1
''' % (rlow, rhigh)
print query
#read data
dt = {}
for x in data[0]:
dt[x] = N.loadtxt(data[0][x])
dt['Bolshoi'] = db.sqlite.get_data_sqlite(path, database, query) * 1e9
#calculate the mass functions from the Bolshoi data
mbin0, mf0 = df.diffFunctionLogBinning(dt['Bolshoi'] / h,
nbins=35,
h=0.7,
mmin=10 ** 9.0,
mmax=10 ** 15.0,
volume=50,
nvols=26,
physical_units=True)
del dt['Bolshoi']
#use chain rule to get dN / dM
#dN/dM = dN/dlog10(M) * dlog10(M)/dM
#d/dM (log10(M)) = 1 / (M*ln(10))
mf0 *= 1. / (mbin0 * N.log(10))
mbin0 = 10 ** mbin0
#title
ax1.set_title('Dark Matter Halo Mass Functions (galpropz.dat)')
#mark redshift
for a, b in zip(mbin0[::-1], mf0[::-1]):
if b > 10 ** -6:
break
ax1.annotate('$z \sim %.0f$' % redshift,
(0.98 * a, 3 * 10 ** -7), size='x-small')
#Analytical MFs
#0th column: log10 of mass (Msolar, NOT Msolar/h)
#1st column: mass (Msolar/h)
#2nd column: (dn/dM)*dM, per Mpc^3 (NOT h^3/Mpc^3)
xST = 10 ** dt['Sheth-Tormen'][:, 0]
yST = dt['Sheth-Tormen'][:, 2] * fudge
sh = ax1.plot(xST, yST, 'b-', lw=1.3)
#PS
xPS = 10 ** dt['Press-Schecter'][:, 0]
yPS = dt['Press-Schecter'][:, 2] * fudge
ps = ax1.plot(xPS, yPS, 'g--', lw=1.1)
#MF from Bolshoi
bolshoi = ax1.plot(mbin0, mf0, 'ro:', ms=5)
#delete data to save memory, dt is not needed any longer
del dt
#plot the residuals
if round(float(redshift), 1) < 1.5:
#interploate to right x scale
ySTint = N.interp(mbin0, xST, yST)
yPSint = N.interp(mbin0, xPS, yPS)
#make the plot
ax2.annotate('$z \sim %.0f$' % redshift,
(1.5 * 10 ** 9, 1.05), xycoords='data',
size=10)
ax2.axhline(1.0, color='b')
ax2.plot(mbin0, mf0 / ySTint, 'b-')
ax2.plot(mbin0, mf0 / yPSint, 'g-')
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_ylim(5 * 10 ** -8, 10 ** -1)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(10 ** 9, 10 ** 15)
ax2.set_xlim(10 ** 9, 10 ** 15)
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(r'$\mathrm{d}N / \mathrm{d}M_{\mathrm{vir}} \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$')
ax2.set_ylabel(r'$\frac{\mathrm{galpropz.dat}}{\mathrm{Model}}$')
ax1.legend((bolshoi, sh, ps),
('Bolshoi', 'Sheth-Tormen', 'Press-Schecter'),
shadow=True, fancybox=True,
numpoints=1)
def plot_mass_functionAnalytical2(redshift, h, no_phantoms, *data):
#fudge factor
fudge = 1.
#http://adsabs.harvard.edu/abs/2001MNRAS.321..372J
#Jenkins et al. paper has sqrt(2./pi) while
#Rachel's code has 1/(sqrt(2*pi))
#ratio of these two is the fudge factor
#read data
dt = {}
for x in data[0]:
if 'Bolshoi' in x:
dt[x] = io.readBolshoiDMfile(data[0][x], 0, no_phantoms)
else:
#M dN/dM dNcorr/dM dN/dlog10(M) dN/dlog10(Mcorr)
d = N.loadtxt(data[0][x])
dt['Press-Schecter'] = N.array([d[:, 0], d[:, 3]])
dt['Sheth-Tormen'] = N.array([d[:, 0], d[:, 4]])
#calculate the mass functions from the Bolshoi data
mbin0, mf0 = df.diffFunctionLogBinning(dt['Bolshoi'] / h,
nbins=35,
h=0.7,
mmin=10 ** 9.0,
mmax=10 ** 15.0,
physical_units=True)
del dt['Bolshoi']
mbin0 = 10 ** mbin0
#title
if no_phantoms:
ax1.set_title('Bolshoi Dark Matter Mass Functions (no phantoms)')
else:
ax1.set_title('Bolshoi Dark Matter Mass Functions')
#mark redshift
for a, b in zip(mbin0[::-1], mf0[::-1]):
if b > 10 ** -6:
break
ax1.annotate('$z \sim %.1f$' % redshift,
(0.98 * a, 3 * 10 ** -6), size='x-small')
#Analytical MFs
xST = dt['Sheth-Tormen'][0]
yST = dt['Sheth-Tormen'][1] * fudge
print xST[1000], yST[1000]
sh = ax1.plot(xST, yST, 'b-', lw=1.3)
#PS
xPS = dt['Press-Schecter'][0]
yPS = dt['Press-Schecter'][1] * fudge
ps = ax1.plot(xPS, yPS, 'g--', lw=1.1)
#MF from Bolshoi
bolshoi = ax1.plot(mbin0, mf0, 'ro:', ms=5)
#delete data to save memory, dt is not needed any longer
del dt
#plot the residuals
if round(float(redshift), 1) < 1.5:
#interploate to right x scale
mfintST = N.interp(xST, mbin0, mf0)
mfintPS = N.interp(xPS, mbin0, mf0)
#yST = N.interp(mbin0, xST, yST)
#yPS = N.interp(mbin0, xPS, yPS)
#make the plot
ax2.annotate('$z \sim %.0f$' % redshift,
(1.5 * 10 ** 9, 1.05), xycoords='data',
size=10)
ax2.axhline(1.0, color='b')
ax2.plot(xST, mfintST / yST, 'b-')
ax2.plot(xPS, mfintPS / yPS, 'g-')
#ax2.plot(mbin0, mf0 / yST, 'b-')
#ax2.plot(mbin0, mf0 / yPS, 'g-')
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_ylim(3 * 10 ** -7, 10 ** 0)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(10 ** 9, 10 ** 15)
ax2.set_xlim(10 ** 9, 10 ** 15)
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(r'$\mathrm{d}N / \mathrm{d}\log_{10}(M_{\mathrm{vir}}) \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$')
ax2.set_ylabel(r'$\frac{\mathrm{Bolshoi}}{\mathrm{Model}}$')
ax1.legend((bolshoi, sh, ps),
('Bolshoi', 'Sheth-Tormen', 'Press-Schecter'),
shadow=True, fancybox=True,
numpoints=1)
def plotMassFunction(simus, sheth, outdir, h, simuPath, simuDB,
noPhantoms=True):
'''
This function plots dark matter halo mass functions from
file(s) given in simus. The usual usage is that simus
contain a list of file names. These files should contain
dark matter halo masses derived from the Bolshoi isotree
files.
This function uses a fudge factor = 2.0 because
Jenkins et al. paper
(http://adsabs.harvard.edu/abs/2001MNRAS.321..372J)
has sqrt(2./pi) while
Rachel's code has 1/(sqrt(2*pi))
The ratio of these two definitions is the fudge factor.
'''
#fudge factor
fudge = 2.0
#figure definitions
left, width = 0.1, 0.8
rect1 = [left, 0.1, width, 0.2]
rect2 = [left, 0.3, width, 0.65]
for i, file in enumerate(simus):
#get the redshift
redshift = float(file.split('z')[1][:5])
logging.debug('Plotting redshift %.2f dark matter mass functions', redshift)
#different redshift ranges
if redshift < 1.8:
rlow = redshift - 0.02
rhigh = redshift + 0.02
else:
rlow = redshift - 0.1
rhigh = redshift + 0.1
#read bolshoi data
BoshoiData = io.readBolshoiDMfile(file, 0, noPhantoms)
#calculate the mass functions from the Bolshoi data
mbin0, mf0 = df.diffFunctionLogBinning(BoshoiData / h,
nbins=35,
h=h,
mmin=1e9,
mmax=1e15,
volume=50.0,
nvols=18)
#use chain rule to get dN / dM
#dN/dM = dN/dlog10(M) * dlog10(M)/dM
#d/dM (log10(M)) = 1 / (M*ln(10))
mf0 *= 1. / (mbin0 * N.log(10))
#put mass back to power
mbin0 = 10 ** mbin0
#get the SAMs data
#if round(redshift, 2) < 0.08:
if i < 1:
query = '''select mhalo from galprop where galprop.gal_id = 1'''
GalpropData = 10 ** db.sqlite.get_data_sqlite(simuPath, simuDB, query)
else:
query = '''select mhalo from galpropz where
galpropz.zgal > {0:f} and galpropz.zgal < {1:f} and
galpropz.gal_id = 1'''.format(rlow, rhigh)
GalpropData = db.sqlite.get_data_sqlite(simuPath, simuDB, query) * 1e9
logging.debug(query)
#calculate the mass functions from the SAM data, only x volumes
mbin0SAM, mf0SAM = df.diffFunctionLogBinning(GalpropData,
nbins=35,
h=h,
mmin=1e9,
mmax=1e15,
volume=50.0,
nvols=15)
#use chain rule to get dN / dM
#dN/dM = dN/dlog10(M) * dlog10(M)/dM
#d/dM (log10(M)) = 1 / (M*ln(10))
mf0SAM *= 1. / (mbin0SAM * N.log(10))
mbin0SAM = 10 ** mbin0SAM
#start a figure
fig = P.figure(figsize=(10, 10))
ax1 = fig.add_axes(rect2) #left, bottom, width, height
ax2 = fig.add_axes(rect1)
#title
if noPhantoms:
ax1.set_title('$z \sim %.2f$ (no phantoms)' % redshift)
else:
ax1.set_title('$z \sim %.2f$' % redshift)
#Analytical MFs
#0th column: log10 of mass (Msolar, NOT Msolar/h)
#1st column: mass (Msolar/h)
#2nd column: (dn/dM)*dM, per Mpc^3 (NOT h^3/Mpc^3)
logging.debug(sheth[i])
dt = N.loadtxt(sheth[i])
xST = 10 ** dt[:, 0]
yST = dt[:, 2] * fudge
sh = ax1.plot(xST, yST, 'b-', lw=1.3)
#MF from Bolshoi
bolshoi = ax1.plot(mbin0, mf0, 'ro:', ms=5)
#MF from the SAM run
samax = ax1.plot(mbin0SAM, mf0SAM, 'gs--', ms=4)
#plot the residuals
ax2.axhline(1.0, color='k')
ax2.plot(mbin0, mf0 / mf0SAM, 'm-')
#interploate to right x scale
mfintST1 = N.interp(xST, mbin0, mf0)
mfintST2 = N.interp(xST, mbin0SAM, mf0SAM)
ax2.plot(xST, mfintST1 / yST, 'r-')
ax2.plot(xST, mfintST2 / yST, 'g-')
ax1.set_ylim(1e-7, 1e-1)
ax2.set_ylim(0.45, 1.55)
ax1.set_xlim(1e9, 1e15)
ax2.set_xlim(1e9, 1e15)
ax1.set_xscale('log')
ax2.set_xscale('log')
ax1.set_yscale('log')
ax1.set_xticklabels([])
ax2.set_xlabel(r'$M_{\mathrm{vir}} \quad [M_{\odot}]$')
ax1.set_ylabel(r'$\mathrm{d}N / \mathrm{d}M_{\mathrm{vir}} \quad [\mathrm{Mpc}^{-3} \mathrm{dex}^{-1}]$')
ax2.set_ylabel(r'$\frac{\mathrm{Bolshoi}}{\mathrm{Model}}$')
ax1.legend((bolshoi, samax, sh),
('Bolshoi', 'galpropz', 'Sheth-Tormen'),
shadow=True, fancybox=True, numpoints=1)
P.savefig(outdir + 'DMMF{0:d}.png'.format(i))
P.close()
