xx = np.linspace(0, M, N)
f = 0.0
for x1 in xx:
f += arnaud_profile_2d(x, x1, Rs, M500, R500, zi, rho_crit, hz, omega_b0, omega_m0, cosmo_h)
#print xx
f *= (2 * (xx[1] - xx[0]))
return f
if __name__=='__main__':
from scipy.interpolate import interp1d
z = 1. #0.0231
cosmo = CosmologyFunctions(z, 'wlsz.ini', 'battaglia')
omega_b0 = cosmo._omega_b0
omega_m0 = cosmo._omega_m0
cosmo_h = cosmo._h
BryanDelta = cosmo.BryanDelta()
rho_critical = cosmo.rho_crit() * cosmo._h * cosmo._h
rarr = np.logspace(-3, 3, 100)
Mvir = 1.e15 #/ cosmo_h
Mvir, Rvir, M200, R200, rho_s, Rs = MvirToMRfrac(Mvir, z, BryanDelta, rho_critical, cosmo_h)
print('%.2e %.2f %.2e %.2f %.2e %.2f'%(Mvir, Rvir, M200, R200, rho_s, Rs))
M200 = 8.915e14
R200 = 1.392
Rs = 0.53
xarr = rarr / Rs
pe_ba = np.array([battaglia_profile_2d(x, 0., Rs, M200, R200, z, rho_critical, omega_b0, omega_m0, cosmo_h) for x in xarr])
pl.subplot(121)
pl.loglog(np.logspace(-3, 3, 100), pe_ba, label='Vinu')
def cl_WL_tSZ(config_file, cosmology, fwhm_k, fwhm_y, kk, yy, ky, zsfile,
P01, P02, P03, xc1, xc2, xc3, beta1, beta2, beta3,
omega_m0, sigma_8, odir, ofile):
'''
Compute WL X tSZ halomodel for a given source redshift distribution
'''
if ky:
sigma_k = fwhm_k * np.pi / 2.355 / 60. /180. #angle in radian
sigma_y = fwhm_y * np.pi / 2.355 / 60. /180. #angle in radian
sigmasq = sigma_k * sigma_y
elif kk:
sigma_k = fwhm_k * np.pi / 2.355 / 60. /180. #angle in radian
sigmasq = sigma_k * sigma_k
elif yy:
sigma_y = fwhm_y * np.pi / 2.355 / 60. /180. #angle in radian
sigmasq = sigma_y * sigma_y
else:
raise ValueError('Either kk, yy or ky should be True')
config = ConfigParser()
config.read(config_file)
if omega_m0 is None:
omega_m0 = config.getfloat(cosmology, 'omega_m0')
if sigma_8 is None:
sigma_8 = config.getfloat(cosmology, 'sigma_8')
omega_l0 = 1. - omega_m0
omega_b0 = config.getfloat(cosmology, 'omega_b0')
h = config.getfloat(cosmology, 'h')
n_scalar = config.getfloat(cosmology, 'n_scalar')
cosmo_dict = {'omega_m0':omega_m0, 'omega_l0':omega_l0, 'omega_b0':omega_b0, 'h':h, 'sigma_8':sigma_8, 'n_scalar':n_scalar}
cosmo0 = CosmologyFunctions(0, config_file, cosmo_dict)
cosmo_h = cosmo0._h
save_clfile = config.getboolean('halomodel', 'save_clfile')
print_cl = config.getboolean('halomodel', 'print_cl')
light_speed = config.getfloat('halomodel', 'light_speed') #km/s
mpctocm = config.getfloat('halomodel', 'mpctocm')
kB_kev_K = config.getfloat('halomodel', 'kB_kev_K')
sigma_t_cm = config.getfloat('halomodel', 'sigma_t_cm') #cm^2
rest_electron_kev = config.getfloat('halomodel', 'rest_electron_kev') #keV
constk = 3. * omega_m0 * (cosmo_h * 100. / light_speed)**2. / 2. #Mpc^-2
consty = mpctocm * sigma_t_cm / rest_electron_kev
zsarr, Ns = np.genfromtxt(zsfile, unpack=True)
if np.isscalar(zsarr):
zsarr = np.array([zsarr])
Ns = np.array([Ns])
else:
zint = np.sum(Ns) * (zsarr[1] - zsarr[0])
Ns /= zint
if P01 is None:
P01 = config.getfloat('halomodel', 'P01')
if P02 is None:
P02 = config.getfloat('halomodel', 'P02')
if P03 is None:
P03 = config.getfloat('halomodel', 'P03')
if xc1 is None:
xc1 = config.getfloat('halomodel', 'xc1')
if xc2 is None:
xc2 = config.getfloat('halomodel', 'xc2')
if xc3 is None:
xc3 = config.getfloat('halomodel', 'xc3')
if beta1 is None:
beta1 = config.getfloat('halomodel', 'beta1')
if beta2 is None:
beta2 = config.getfloat('halomodel', 'beta2')
if beta3 is None:
beta3 = config.getfloat('halomodel', 'beta3')
kRmax = config.getfloat('halomodel', 'kRmax')
kRspace = config.getint('halomodel', 'kRspace')
yRmax = config.getfloat('halomodel', 'yRmax')
yRspace = config.getint('halomodel', 'yRspace')
kmin = config.getfloat('halomodel', 'kmin') #1/Mpc
kmax = config.getfloat('halomodel', 'kmax')
kspace = config.getint('halomodel', 'kspace')
ellmin = config.getint('halomodel', 'ellmin')
ellmax = config.getfloat('halomodel', 'ellmax')
ellspace = config.getint('halomodel', 'ellspace')
mmin = config.getfloat('halomodel', 'mmin')
mmax = config.getfloat('halomodel', 'mmax')
mspace = config.getfloat('halomodel', 'mspace')
zmin = config.getfloat('halomodel', 'zmin')
zmax = config.getfloat('halomodel', 'zmax')
zspace = config.getfloat('halomodel', 'zspace')
dlnk = np.log(kmax/kmin) / kspace
lnkarr = np.linspace(np.log(kmin), np.log(kmax), kspace)
karr = np.exp(lnkarr).astype(np.float64)
#No little h
#Input Mpc/h to power spectra and get Mpc^3/h^3
pk_arr = np.array([cosmo0.linear_power(k/cosmo0._h) for k in karr]).astype(np.float64) / cosmo0._h / cosmo0._h / cosmo0._h
#np.savetxt('pk_vinu.dat', np.transpose((karr/cosmo0._h, pk_arr*cosmo0._h**3)))
#sys.exit()
#kc, pkc = np.genfromtxt('matterpower.dat', unpack=True)
#pkspl = InterpolatedUnivariateSpline(kc, pkc, k=2)
#pkh_arr = pkspl(karr/cosmo0._h)
#pk_arr = pkh_arr / cosmo0._h / cosmo0._h / cosmo0._h
pkspl = InterpolatedUnivariateSpline(karr, pk_arr, k=2)
#pl.loglog(karr, pk_arr)
#pl.show()
dlnm = np.log(mmax/mmin) / mspace
lnmarr = np.linspace(np.log(mmin), np.log(mmax), mspace)
marr = np.exp(lnmarr).astype(np.float64)
lnzarr = np.linspace(np.log(1.+zmin), np.log(1.+zmax), zspace)
zarr = np.exp(lnzarr) - 1.0
dlnz = np.log((1.+zmax)/(1.+zmin)) / zspace
print('dlnk, dlnm dlnz', dlnk, dlnm, dlnz)
if doPrintCl:
print('dlnk, dlnm dlnz', dlnk, dlnm, dlnz)
#No little h
#Need to give mass * h and get the sigma without little h
#The following lines are used only used for ST MF and ST bias
sigma_m0 = np.array([cosmo0.sigma_m(m * cosmo0._h) for m in marr])
rho_norm0 = cosmo0.rho_bar()
lnMassSigmaSpl = InterpolatedUnivariateSpline(lnmarr, sigma_m0, k=3)
hzarr, BDarr, rhobarr, chiarr, dVdzdOm, rho_crit_arr = [], [], [], [], [], []
bias, bias_t, Darr = [], [], []
marr2, mf, dlnmdlnm, Deltaarr = [], [], [], []
if config.get('halomodel', 'MF') == 'Tinker' and config.get('halomodel', 'MassToIntegrate') == 'm200m':
#Mass using critical density (ie. m200c)
tm200c = np.logspace(np.log10(1e8), np.log10(1e17), 50)
#m200m mass using mean mass density
tmarr = np.exp(lnmarr).astype(np.float64)
#tf = np.genfromtxt('../data/z_m_relation.dat')
#tz = tf[:,0]
#tmv = tf[:,1]
#tm200c = tf[:,2]
#tm200m = tf[:,3]
for i, zi in enumerate(zarr):
cosmo = CosmologyFunctions(zi, config_file, cosmo_dict)
rcrit = cosmo.rho_crit() * cosmo._h * cosmo._h
rbar = cosmo.rho_bar() * cosmo._h * cosmo._h
bn = cosmo.BryanDelta()
BDarr.append(bn) #OK
rho_crit_arr.append(rcrit) #OK
rhobarr.append(rbar)
chiarr.append(cosmo.comoving_distance() / cosmo._h)
hzarr.append(cosmo.E0(zi))
DD = bn/cosmo.omega_m()
DD200 = 200./cosmo.omega_m()
#Number of Msun objects/Mpc^3 (i.e. unit is 1/Mpc^3)
if config.get('halomodel', 'MF') == 'Tinker':
if config.get('halomodel', 'MassToIntegrate') == 'virial':
if DD > 200:
mFrac = marr * cosmo_h
print(1)
mFrac = marr * h0
#print bn, cosmo.omega_m(), bn/cosmo.omega_m()
mf.append(bias_mass_func_tinker(zi, config_file, cosmo_dict, mFrac.min(), mFrac.max(), mspace, bias=False, Delta=DD, marr=mFrac, reduced=False)[1])
marr2.append(marr)
dlnmdlnm.append(np.ones_like(marr))
else:
mFrac = np.array([HuKravtsov(zi, mv, rcrit, bn, config.getfloat('halomodel', 'MassDef') * cosmo.omega_m(), cosmo_h, 1)[2] for mv in marr]) * cosmo_h
mf.append(bias_mass_func_tinker(zi, config_file, cosmo_dict, mFrac.min(), mFrac.max(), mspace, bias=False, Delta=config.getfloat('halomodel', 'MassDef'), marr=mFrac, reduced=False)[1])
marr2.append(marr)
dlnmdlnm.append([dlnMdensitydlnMcritOR200(config.getfloat('halomodel', 'MassDef') * cosmo.omega_m(), bn, mFm/cosmo_h, mv, zi, cosmo_h, 1) for mv,mFm in zip(marr, mFrac)]) #dlnmFrac/dlnMv. In the bias_mass_func_tinker() I have computed dn/dlnM where M is in the unit of Msol. I have therefore include h in that mass function. Therefore, I just need to multiply dlnmFrac/dlnMv only
print(2)
mFrac = np.array([HuKravtsov(zi, mv, rcrit, bn, float(config['massfunc']['MassDef'])/cosmo.omega_m(), h0, 1)[2] for mv in marr]) * h0
mf.append(bias_mass_func_tinker(zi, mFrac.min(), mFrac.max(), mspace, bias=False, Delta=float(config['massfunc']['MassDef']), marr=mFrac, reduced=False)[1])
marr2.append(marr)
dlnmdlnm.append([dlnMdensitydlnMcritOR200(float(config['massfunc']['MassDef']) / cosmo.omega_m(), bn, mFm/h0, mv, zi, h0, 1) for mv,mFm in zip(marr, mFrac)]) #dlnmFrac/dlnMv. In the bias_mass_func_tinker() I have computed dn/dlnM where M is in the unit of Msol. I have therefore include h in that mass function. Therefore, I just need to multiply dlnmFrac/dlnMv only
#print dlnmdlnm
#print a
input_mvir = 1
elif config.get('halomodel', 'MassToIntegrate') == 'm200c':
#XXX
#m200m = np.array([HuKravtsov(zi, mv, rcrit, 200, 200*cosmo.omega_m(), cosmo_h, 0)[2] for mv in marr]) #* cosmo_h
#print m200m
#XXX
if DD200 > 200:
mFrac = marr * cosmo_h
#print 200, cosmo.omega_m(), 200/cosmo.omega_m()
mf.append(bias_mass_func_tinker(zi, config_file, cosmo_dict, mFrac.min(), mFrac.max(), mspace, bias=False, Delta=DD200, marr=mFrac, reduced=False)[1])
marr2.append(marr)
dlnmdlnm.append(np.ones_like(marr))
else:
mFrac = np.array([HuKravtsov(zi, m2c, rcrit, 200, config.getfloat('halomodel', 'MassDef') * cosmo.omega_m(), cosmo_h, 0)[2] for m2c in marr]) * cosmo_h
mf.append(bias_mass_func_tinker(zi, config_file, cosmo_dict, mFrac.min(), mFrac.max(), mspace, bias=False, Delta=config.getfloat('halomodel', 'MassDef'), marr=mFrac)[1])
marr2.append(marr)
for m2,mFm in zip(marr, mFrac):
dlnmdlnm.append(dlnMdensitydlnMcritOR200(config.getfloat('halomodel', 'MassDef') * cosmo.omega_m(), 200., mFm/cosmo_h, m2, zi, cosmo_h, 0)) #dlnM200m/dlnMv. In the bias_mass_func_tinker() I have computed dn/dlnM where M is in the unit of Msol. I have therefore include h in that mass function. Therefore, I just need to multiply dlnM200m/dlnMv only
mFrac = np.array([HuKravtsov(zi, m2c, rcrit, 200, float(config['massfunc']['MassDef'])/cosmo.omega_m(), h0, 0)[2] for m2c in marr]) * h0
mf.append(bias_mass_func_tinker(zi, mFrac.min(), mFrac.max(), mspace, bias=False, Delta=float(config['massfunc']['MassDef']), marr=mFrac)[1])
marr2.append(marr)
for m2,mFm in zip(marr, mFrac):
dlnmdlnm.append(dlnMdensitydlnMcritOR200(float(config['massfunc']['MassDef']) / cosmo.omega_m(), 200., mFm/h0, m2, zi, h0, 0)) #dlnM200m/dlnMv. In the bias_mass_func_tinker() I have computed dn/dlnM where M is in the unit of Msol. I have therefore include h in that mass function. Therefore, I just need to multiply dlnM200m/dlnMv only
input_mvir = 0
elif config.get('halomodel', 'MassToIntegrate') == 'm200m':
#raise ValueError('Use MassToIntegrate=virial/m200c. m200m is not working')
#Temporary mass array of m200m from m200c
tm200m = np.array([HuKravtsov(zi, tt, rcrit, 200, 200.*cosmo.omega_m(), cosmo._h, 0)[2] for tt in tm200c])
#m200m vs m200c spline
tmspl = InterpolatedUnivariateSpline(tm200m, tm200c)
#Now m200c from m200m, i.e. tmarr which is the integrating
#variable
m200c = tmspl(tmarr)
#m200m Msol/h
m200m = tmarr * cosmo_h
marr2.append(m200c)
mf.append(bias_mass_func_tinker(zi, config_file, cosmo_dict, m200m.min(), m200m.max(), mspace, bias=False, Delta=200, marr=m200m)[1])
input_mvir = 0
elif config.get('halomodel', 'MF') == 'Bocquet':
if config.get('halomodel', 'MassToIntegrate') == 'virial':
m200 = np.array([HuKravtsov(zi, mv, rcrit, bn, 200, cosmo_h, 1)[2] for mv in marr])
mf.append(bias_mass_func_bocquet(zi, config_file, cosmo_dict, m200.min(), m200.max(), mspace, bias=False, marr=m200)[1])
marr2.append(marr)
for mv,m2 in zip(marr, m200):
dlnmdlnm.append(dlnMdensitydlnMcritOR200(200., bn, m2, mv, zi, cosmo_h, 1))
input_mvir = 1
elif config.get('halomodel', 'MassToIntegrate') == 'm200':
tmf = bias_mass_func_bocquet(zi, config_file, cosmo_dict, marr.min(), marr.max(), mspace, bias=False, marr=marr)[1]
mf.append(tmf)
dlnmdlnm.append(np.ones(len(tmf)))
input_mvir = 0
elif config.get('halomodel', 'MF') == 'ST':
mf.append(bias_mass_func_st(zi, config_file, cosmo_dict, marr.min(), marr.max(), mspace, bias=False, marr=marr)[1])
marr2.append(marr)
dlnmdlnm.append(np.ones_like(marr))
input_mvir = 1
#raise ValueError('MF should be Tinker or Bocquet')
#Bias is calculated by assuming that the mass is virial. I need to change that
#print DD
#for m in marr:
# print '%.2e T %.2f'%(m, halo_bias_tinker(DD, cosmo.delta_c() / cosmo._growth / lnMassSigmaSpl(np.log(m))))
# print '%.2e ST %.2f'%(m, halo_bias_st(cosmo.delta_c() * cosmo.delta_c() / cosmo._growth / cosmo._growth / lnMassSigmaSpl(np.log(m)) / lnMassSigmaSpl(np.log(m))))
#sys.exit()
if config.get('halomodel', 'MF') == 'Tinker':
if DD >= 200:
bias.append(np.array([halo_bias_tinker(DD, cosmo.delta_c() / cosmo._growth / lnMassSigmaSpl(np.log(m))) for m in marr]))
else:
bias.append(np.array([halo_bias_tinker(config.getfloat('halomodel', 'MassDef') * cosmo.omega_m(), cosmo.delta_c() / cosmo._growth / lnMassSigmaSpl(np.log(m))) for m in mFrac/cosmo_h]))
elif config.get('halomodel', 'MF') == 'ST' or config.get('halomodel', 'MF') == 'Bocquet':
bias.append(np.array([halo_bias_st(cosmo.delta_c() * cosmo.delta_c() / cosmo._growth / cosmo._growth / lnMassSigmaSpl(np.log(m)) / lnMassSigmaSpl(np.log(m))) for m in marr]))
dVdzdOm.append(cosmo.E(zi) / cosmo._h) #Mpc/h, It should have (km/s/Mpc)^-1 but in the cosmology code the speed of light is removed
Darr.append(cosmo._growth)
#pl.title('%.2f'%zi)
#pl.scatter(np.array([halo_bias_st(cosmo.delta_c() * cosmo.delta_c() / cosmo._growth / cosmo._growth / lnMassSigmaSpl(np.log(m)) / lnMassSigmaSpl(np.log(m))) for m in marr]), np.array([halo_bias_tinker(200, cosmo.delta_c() / cosmo._growth / lnMassSigmaSpl(np.log(m))) for m in marr]))
#pl.show()
if config.get('halomodel', 'MF') =='Tinker' and config.get('halomodel', 'MassToIntegrate') == 'm200m':
mf = np.array(mf).flatten()
else:
mf = np.array(mf).flatten() * np.array(dlnmdlnm).flatten()
hzarr = np.array(hzarr)
BDarr = np.array(BDarr)
rhobarr = np.array(rhobarr)
chiarr = np.array(chiarr)
dVdzdOm = np.array(dVdzdOm) * chiarr * chiarr
rho_crit_arr = np.array(rho_crit_arr)
marr2 = np.array(marr2).flatten()
zchispl = InterpolatedUnivariateSpline(zarr, chiarr, k=2)
chisarr = zchispl(zsarr)
bias = np.array(bias).flatten()
bias_t = np.array(bias_t).flatten()
Darr = np.array(Darr)
#ellarr = np.linspace(1, 10001, 10)
ellarr = np.logspace(np.log10(ellmin), np.log10(ellmax), ellspace)
cl_arr, cl1h_arr, cl2h_arr = [], [], []
for ell in ellarr:
pk = pkspl(ell/chiarr)
if ky:
cl1h, cl2h, cl = integrate_kyhalo(ell, lnzarr, chiarr, dVdzdOm, marr2, mf, BDarr, rhobarr, rho_crit_arr, bias, Darr, pk, zsarr, chisarr, Ns, dlnz, dlnm, omega_b0, omega_m0, cosmo_h, constk, consty, input_mvir, mspace, kRmax, kRspace, yRmax, yRspace, P01, P02, P03, xc1, xc2, xc3, beta1, beta2, beta3)
if kk:
cl1h, cl2h, cl = integrate_kkhalo(ell, lnzarr, chiarr, dVdzdOm, marr2, mf, BDarr, rhobarr, rho_crit_arr, bias, Darr, pk, zsarr, chisarr, Ns, dlnz, dlnm, omega_b0, omega_m0, cosmo_h, constk, consty, input_mvir, mspace, kRmax, kRspace)
if yy:
cl1h, cl2h, cl = integrate_yyhalo(ell, lnzarr, chiarr, dVdzdOm, marr2, mf, BDarr, rhobarr, rho_crit_arr, bias, Darr, pk, dlnz, dlnm, omega_b0, omega_m0, cosmo_h, constk, consty, input_mvir, mspace, yRmax, yRspace, P01, P02, P03, xc1, xc2, xc3, beta1, beta2, beta3)
cl_arr.append(cl)
cl1h_arr.append(cl1h)
cl2h_arr.append(cl2h)
if print_cl:
print('l %.2 Cl_1h %.2e Cl_2h %.2e Cl %.2e'%(ell, cl1h, cl2h, cl))
if doPrintCl:
print(ell, cl1h, cl2h, cl)
convolve = np.exp(-1 * sigmasq * ellarr * ellarr)# i.e. the output is Cl by convolving by exp(-sigma^2 l^2)
cl = np.array(cl_arr) * convolve
cl1h = np.array(cl1h_arr) * convolve
cl2h = np.array(cl2h_arr) * convolve
if save_clfile:
np.savetxt(os.path.join(odir, ofile), np.transpose((ellarr, cl1h, cl2h, cl)), fmt='%.2f %.3e %.3e %.3e', header='l Cl1h Cl2h Cl')
return ellarr, cl1h, cl2h, cl