def plot_luminosityfunctionPaper(path,
database,
redshifts,
bands,
out_folder,
solid_angle=100 * 160.,
ymin=5e-7,
ymax=5 * 10**-2,
xmin=9.1,
xmax=13.1,
H0=70.0,
WM=0.28):
"""
:param solid_angle: area of the sky survey in arcmin**2
GOODS = 160, hence 100*160
"""
col = ['black', 'red', 'magenta', 'green', 'blue']
lin = [':', '--', '-', '-.', '-']
#make the figure
fig = P.figure()
fig.subplots_adjust(left=0.09,
bottom=0.08,
right=0.93,
top=0.95,
wspace=0.0,
hspace=0.0)
ax1 = fig.add_subplot(221)
ax2 = fig.add_subplot(222)
ax3 = fig.add_subplot(223)
ax4 = fig.add_subplot(224)
for b in bands:
if '100' in b: nb = 19
if '160' in b: nb = 14
if '250' in b: nb = 14
if '350' in b: nb = 11
print '\nPlotting ', b
#redshift limited plots
for i, red in enumerate(redshifts):
query = '''select %s from FIR where %s > 7.7 and %s
and FIR.spire250_obs < 1e6''' % (b, b, red)
limited = SamPy.db.sqlite.get_data_sqlite(path, database, query)
print query, len(limited)
#modify redshift string
tmp = red.split()
#rtitle = r'$z = %.0f$' % N.mean(N.array([float(tmp[2]), float(tmp[6])]))
rtitle = r'$%s \leq z < %s$' % (tmp[2], tmp[6])
#get a comoving volume
comovingVol = cv.comovingVolume(solid_angle,
float(tmp[2]),
float(tmp[6]),
H0=H0,
WM=WM)
#weights
wghts = N.zeros(len(limited)) + (1. / comovingVol)
#differential function
bb, nn, nu = df.diff_function_log_binning(limited,
wgth=wghts,
mmax=xmax,
mmin=xmin,
nbins=nb,
log=True)
mask = nu > 0
x = bb[mask]
y = nn[mask]
#to make sure that the plots go below the area plotted
x = N.append(x, N.max(x) * 1.01)
y = N.append(y, 1e-10)
if '100' in b:
ax1.plot(x,
N.log10(y),
color=col[i],
marker='None',
ls=lin[i],
label=rtitle)
if '160' in b:
ax2.plot(x,
N.log10(y),
color=col[i],
marker='None',
ls=lin[i],
label=rtitle)
if '250' in b:
ax3.plot(x,
N.log10(y),
color=col[i],
marker='None',
ls=lin[i],
label=rtitle)
if '350' in b:
ax4.plot(x,
N.log10(y),
color=col[i],
marker='None',
ls=lin[i],
label=rtitle)
#plot observational constrains
mic100 = lf.Herschel100Lapi()
mic250 = lf.Herschel250Lapi()
#x values
x100 = N.log10(mic100['Lsun'])
x250 = N.log10(mic250['Lsun'])
#mask out missing values
msk100z15 = mic100['z1.5'][:, 0] > -6.5
msk100z22 = mic100['z2.2'][:, 0] > -6.5
msk100z32 = mic100['z3.2'][:, 0] > -6.5
msk250z15 = mic250['z1.4'][:, 0] > -6.5
msk250z22 = mic250['z2.2'][:, 0] > -6.5
msk250z32 = mic250['z3.2'][:, 0] > -6.5
#PACS100 plots
ax1.errorbar(x100[msk100z15],
mic100['z1.5'][:, 0][msk100z15],
yerr=[
-mic100['z1.5'][:, 2][msk100z15],
mic100['z1.5'][:, 1][msk100z15]
],
marker='s',
ms=4,
ls='None',
mfc='r',
mec='r',
c='r')
ax1.errorbar(x100[msk100z22],
mic100['z2.2'][:, 0][msk100z22],
yerr=[
-mic100['z2.2'][:, 2][msk100z22],
mic100['z2.2'][:, 1][msk100z22]
],
marker='o',
ms=4,
ls='None',
mfc='m',
mec='m',
c='m')
ax1.errorbar(x100[msk100z32],
mic100['z3.2'][:, 0][msk100z32],
yerr=[
-mic100['z3.2'][:, 2][msk100z32],
mic100['z3.2'][:, 1][msk100z32]
],
marker='d',
ms=4,
ls='None',
mfc='g',
mec='g',
c='g')
# ax1.scatter(x100[msk100z15], 10**mic100['z1.5'][:,0][msk100z15],
# marker='s', s=10,c='k')
# ax1.scatter(x100[msk100z22], 10**mic100['z2.2'][:,0][msk100z22],
# marker='o', s=10, c='r')
# ax1.scatter(x100[msk100z32], 10**mic100['z3.2'][:,0][msk100z32],
# marker='d', s=10, c='m')
#SPIRE250 plots
ax3.errorbar(x250[msk250z15],
mic250['z1.4'][:, 0][msk250z15],
yerr=[
-mic250['z1.4'][:, 2][msk250z15],
mic250['z1.4'][:, 1][msk250z15]
],
marker='s',
ms=4,
ls='None',
mfc='r',
mec='r',
c='r',
label=r'$1.2 \leq z < 1.6$')
ax3.errorbar(x250[msk250z22],
mic250['z2.2'][:, 0][msk250z22],
yerr=[
-mic250['z2.2'][:, 2][msk250z22],
mic250['z2.2'][:, 1][msk250z22]
],
marker='o',
ms=4,
ls='None',
mfc='m',
mec='m',
c='m',
label=r'$2.0 \leq z < 2.4$')
ax3.errorbar(x250[msk250z32],
mic250['z3.2'][:, 0][msk250z32],
yerr=[
-mic250['z3.2'][:, 2][msk250z32],
mic250['z3.2'][:, 1][msk250z32]
],
marker='d',
ms=4,
ls='None',
mfc='g',
mec='g',
c='g',
label=r'$2.4 \leq z < 4.0$')
# ax3.scatter(x250[msk250z15], 10**mic250['z1.4'][:,0][msk250z15],
# marker='s', s=10,c='k')
# ax3.scatter(x250[msk250z22], 10**mic250['z2.2'][:,0][msk250z22],
# marker='o', s=10, c='r')
# ax3.scatter(x250[msk250z32], 10**mic250['z3.2'][:,0][msk250z32],
# marker='d', s=10, c='m')
#labels
ax4.errorbar([
5,
], [
-10,
],
yerr=[
0.1,
],
marker='s',
ms=4,
ls='None',
mfc='r',
mec='r',
c='r',
label=r'$1.2 \leq z < 1.6$')
ax4.errorbar([
5,
], [
-10,
],
yerr=[
0.1,
],
marker='o',
ms=4,
ls='None',
mfc='m',
mec='m',
c='m',
label=r'$2.0 \leq z < 2.4$')
ax4.errorbar([
5,
], [
-10,
],
yerr=[
0.1,
],
marker='d',
ms=4,
ls='None',
mfc='g',
mec='g',
c='g',
label=r'$2.4 \leq z < 4.0$')
#set scales
# ax1.set_yscale('log')
# ax2.set_yscale('log')
# ax3.set_yscale('log')
# ax4.set_yscale('log')
#ylabel = r'$\phi \ [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}]$'
ylabel = r'$\log_{10} \left ( \phi \ [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}] \right )$'
xlabel = r'$\log_{10}(L \ [L_{\odot}])$'
#labels
ax3.set_xlabel(xlabel)
ax4.set_xlabel(xlabel)
ax1.set_ylabel(ylabel)
ax3.set_ylabel(ylabel)
ax2.set_yticklabels([])
ax4.set_yticklabels([])
ax1.set_xticklabels([])
ax2.set_xticklabels([])
#limits
ymin = N.log10(ymin)
ymax = N.log10(ymax)
ax1.set_ylim(ymin, ymax)
ax1.set_xlim(xmin + 0.2, xmax)
ax2.set_ylim(ymin, ymax)
ax2.set_xlim(xmin + 0.2, xmax)
ax3.set_ylim(ymin, ymax)
ax3.set_xlim(xmin + 0.2, xmax)
ax4.set_ylim(ymin, ymax)
ax4.set_xlim(xmin + 0.2, xmax)
#add some annotations
P.text(0.5,
0.94,
'a) PACS 100',
horizontalalignment='center',
verticalalignment='center',
transform=ax1.transAxes)
P.text(0.5,
0.94,
'b) PACS 160',
horizontalalignment='center',
verticalalignment='center',
transform=ax2.transAxes)
P.text(0.5,
0.94,
'c) SPIRE 250',
horizontalalignment='center',
verticalalignment='center',
transform=ax3.transAxes)
P.text(0.5,
0.94,
'd) SPIRE 350',
horizontalalignment='center',
verticalalignment='center',
transform=ax4.transAxes)
#legend
ax4.legend(scatterpoints=1, fancybox=True, shadow=True, loc='center right')
#save figure
P.savefig(out_folder + 'luminosity_functionPaper.ps')
P.close()
def plot_luminosityfunction(path,
database,
redshifts,
band,
out_folder,
solid_angle=10 * 160.,
ymin=10**3,
ymax=2 * 10**6,
xmin=0.5,
xmax=100,
nbins=15,
sigma=5.0,
H0=70.0,
WM=0.28,
zmax=6.0):
"""
:param solid_angle: area of the sky survey in arcmin**2
GOODS = 160, hence 10*160
:param sigma: sigma level of the errors to be plotted
:param nbins: number of bins (for simulated data)
"""
#subplot numbers
columns = 3
rows = 3
#get data
query = """select %s from FIR where %s > 7
and FIR.spire250_obs < 1e6""" % (band, band)
total = SamPy.db.sqlite.get_data_sqlite(path, database, query)
#make the figure
fig = P.figure()
P.subplots_adjust(wspace=0.0, hspace=0.0)
ax = P.subplot(rows, columns, 1)
#get the co-moving volume to the backend
comovingVol = cv.comovingVolume(solid_angle, 0, zmax, H0=H0, WM=WM)
#weight each galaxy
wghts = N.zeros(len(total)) + (1. / comovingVol)
#calculate the differential stellar mass function
#with log binning
b, n, nu = df.diff_function_log_binning(total,
wgth=wghts,
mmax=xmax,
mmin=xmin,
nbins=nbins,
log=True)
#calculate the poisson error
mask = nu > 0
err = wghts[0] * N.sqrt(nu[mask]) * sigma
up = n[mask] + err
lw = n[mask] - err
lw[lw < ymin] = ymin
#plot the sigma area
stot = ax.fill_between(b[mask], up, lw, color='#728FCE')
#plot the knots
mtot = ax.scatter(b[mask], n[mask], marker='o', s=3, color='k')
#add annotation
ax.annotate('Total', (0.5, 0.87), xycoords='axes fraction', ha='center')
#set scale
ax.set_yscale('log')
ax.set_xlim(xmin, xmax)
ax.set_ylim(ymin, ymax)
ax.set_xticklabels([])
ax.set_ylabel(r'$\phi \ [Mpc^{-3} \ dex^{-1}]$')
ptot = P.Rectangle((0, 0), 1, 1, fc='#728FCE')
sline = '%i$\sigma$ errors' % sigma
P.legend((mtot, ptot), ('All Galaxies', sline),
loc='lower left',
scatterpoints=1,
fancybox=True,
shadow=True)
#redshift limited plots
for i, red in enumerate(redshifts):
query = '''select %s from FIR where %s > 7 and %s
and FIR.spire250_obs < 1e6''' % (band, band, red)
limited = SamPy.db.sqlite.get_data_sqlite(path, database, query)
print query, len(limited)
#modify redshift string
tmp = red.split()
#rtitle = r'$z = %.0f$' % N.mean(N.array([float(tmp[2]), float(tmp[6])]))
rtitle = r'$%s < z \leq %s$' % (tmp[2], tmp[6])
#get a comoving volume
comovingVol = cv.comovingVolume(solid_angle,
float(tmp[2]),
float(tmp[6]),
H0=H0,
WM=WM)
#weights
wghts = N.zeros(len(limited)) + (1. / comovingVol)
#differential function
bb, nn, nu = df.diff_function_log_binning(limited,
wgth=wghts,
mmax=xmax,
mmin=xmin,
nbins=nbins,
log=True)
#make a subplot
axs = P.subplot(rows, columns, i + 2)
#calculate the poisson error
mask = nu > 0
err = wghts[0] * N.sqrt(nu[mask]) * sigma
up = nn[mask] + err
lw = nn[mask] - err
lw[lw < ymin] = ymin
#plot the sigma area
axs.fill_between(bb[mask], up, lw, color='#728FCE')
#plot the knots
axs.scatter(bb[mask], nn[mask], marker='o', s=3, color='k')
#add annotation
axs.annotate(rtitle, (0.5, 0.87),
xycoords='axes fraction',
ha='center')
#set scales
axs.set_yscale('log')
axs.set_xlim(xmin, xmax)
axs.set_ylim(ymin, ymax)
#remove unnecessary ticks and add units
if i == 0 or i == 1 or i == 3 or i == 4:
axs.set_yticklabels([])
if i == 2 or i == 3 or i == 4:
btmp = re.search('\d\d\d', band).group()
axs.set_xlabel(r'$\log_{10} (L_{%s} \ [L_{\odot}])$' % btmp)
#axs.set_xticks(axs.get_xticks()[1:])
else:
axs.set_xticklabels([])
if i == 2:
axs.set_ylabel(r'$\phi \ [Mpc^{-3} \ dex^{-1}]$')
#axs.set_xticks(axs.get_xticks()[:-1])
#save figure
P.savefig(out_folder + 'luminosity_function_%s.ps' % band)
P.close()
def plot_luminosityfunctionKDE(path,
database,
redshifts,
band,
out_folder,
solid_angle=10 * 160.,
ymin=10**3,
ymax=2 * 10**6,
xmin=0.5,
xmax=100,
nbins=15,
H0=70.0,
WM=0.28,
zmax=6.0):
"""
:param solid_angle: area of the sky survey in arcmin**2
GOODS = 160, hence 10*160
:param sigma: sigma level of the errors to be plotted
:param nbins: number of bins (for simulated data)
"""
col = ['black', 'red', 'magenta', 'green', 'blue', 'brown']
#get data
query = '''select %s from FIR where %s > 6
and FIR.spire250_obs < 1e6''' % (band, band)
total = SamPy.db.sqlite.get_data_sqlite(path, database, query)[:, 0]
print len(total)
#get the co-moving volume to the backend
comovingVol = cv.comovingVolume(solid_angle, 0, zmax, H0=H0, WM=WM)
#normalization
normalization = float(len(total)) / comovingVol * (nbins * 7 * 2)
#KDE
mu = SS.gaussian_kde(total)
#in which points to evaluate
x = N.linspace(N.min(total), N.max(total), nbins * 7)
#make the figure
fig = P.figure()
ax = P.subplot(111)
#plot
ax.plot(x, mu.evaluate(x) / normalization, color='gray', ls='--')
#redshift limited plots
for i, red in enumerate(redshifts):
query = '''select %s from FIR where %s > 6 and %s
and FIR.spire250_obs < 1e6''' % (band, band, red)
limited = SamPy.db.sqlite.get_data_sqlite(path, database, query)[:, 0]
print query, len(limited)
#modify redshift string
tmp = red.split()
rtitle = r'$%s < z \leq %s$' % (tmp[2], tmp[6])
#get a comoving volume
comovingVol = cv.comovingVolume(solid_angle,
float(tmp[2]),
float(tmp[6]),
H0=H0,
WM=WM)
#normalization
normalization = float(len(limited)) / comovingVol * (nbins * 7 * 2)
#KDE
mu = SS.gaussian_kde(limited)
#in which points to evaluate
x = N.linspace(N.min(limited), N.max(limited), nbins * 7)
ax.plot(x,
mu.evaluate(x) / normalization,
color=col[i],
marker='None',
ls='-',
label=rtitle)
#set scales
ax.set_yscale('log')
ax.set_xlim(xmin, xmax)
ax.set_ylim(ymin, ymax)
ax.set_ylabel(r'$\phi \ [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}]$')
ax.set_xlabel(r'$\log_{10}(L_{%s} \ [L_{\odot}])$' %
re.search('\d\d\d', band).group())
P.legend(scatterpoints=1, fancybox=True, shadow=True)
#save figure
P.savefig(out_folder + 'luminosity_functionKDE_%s.ps' % band)
P.close()
def scatterHistograms(xdata,
ydata,
xlabel,
ylabel,
output):
"""
This functions generates a scatter plot and projected histograms to both axes.
"""
#constants
xmin = 2.0
xmax = 4.0
ymin = -2.05
ymax = 2.5
solid_angle = 100 * 160.
H0 = 70.0
WM = 0.28
#xbins and ybins
xbins = np.linspace(xmin, xmax, 30)
ybins = np.linspace(ymin, ymax, 20)
dfx = xbins[1] - xbins[0]
dfy = ybins[1] - ybins[0]
#calculate volume
comovingVol = cv.comovingVolume(solid_angle,
xmin,
xmax,
H0=H0,
WM=WM)
#weight each galaxy
wghtsx = (np.zeros(len(xdata)) + (1. / comovingVol)) / dfx
wghtsy = (np.zeros(len(ydata)) + (1. / comovingVol)) / dfy
#no labels, null formatter
nullfmt = NullFormatter()
# definitions for the axes
left, width = 0.1, 0.65
bottom, height = 0.1, 0.65
bottom_h = left_h = left + width + 0.02
rect_scatter = [left, bottom, width, height]
rect_histx = [left, bottom_h, width, 0.2]
rect_histy = [left_h, bottom, 0.2, height]
#make a figure
fig = plt.figure(figsize=(12, 12))
axScatter = plt.axes(rect_scatter)
axHistx = plt.axes(rect_histx)
axHisty = plt.axes(rect_histy)
#no labels
axHistx.xaxis.set_major_formatter(nullfmt)
axHisty.yaxis.set_major_formatter(nullfmt)
#the scatter plot
axScatter.scatter(xdata, ydata,
marker='o',
s=0.5,
c='k')
#KDE
x = M.AnaKDE([xdata, ydata])
x_vec, y_vec, zm, lvls, d0, d1 = x.contour(np.linspace(xmin - 0.1, xmax + 0.1, 50),
np.linspace(ymin - 0.1, ymax + 0.1, 50),
return_data=True)
axScatter.contourf(x_vec, y_vec, zm,
levels=np.linspace(0.002, 0.92 * np.max(zm), 10),
cmap=cm.get_cmap('gist_yarg'),
alpha=0.8)
#draw a line to the detection limit, 5mJy (at 250)
axScatter.axhline(np.log10(5),
color='green',
ls='--',
lw=1.8)
#scatter labels
axScatter.set_xlabel(xlabel)
axScatter.set_ylabel(ylabel)
#set scatter limits
axScatter.set_xlim(xmin, xmax)
axScatter.set_ylim(ymin, ymax)
#make x histogram
x1 = axHistx.hist(xdata,
bins=xbins,
weights=wghtsx,
log=True,
color='gray')
x2 = axHistx.hist(xdata[ydata > np.log10(5)],
bins=xbins,
weights=wghtsx[ydata > np.log10(5)],
log=True,
color='gray',
hatch='x')
#set legend of x histogram
plt.legend((x1[2][0], x2[2][0]),
('All Galaxies', r'$S_{160}> 5\ \mathrm{mJy}$'),
shadow=False,
fancybox=False,
bbox_to_anchor=(0.01, 1.34),
loc=2,
borderaxespad=0.)
#make y histogram
axHisty.hist(ydata,
bins=ybins,
orientation='horizontal',
weights=wghtsy,
log=True,
color='gray')
#set histogram limits
axHistx.set_xlim(axScatter.get_xlim())
axHistx.set_ylim(1e-7, 1e-1)
axHisty.set_ylim(axScatter.get_ylim())
axHisty.set_xlim(4e-9, 1e-2)
#set histogram labels
axHistx.set_ylabel(r'$\frac{\mathrm{d}N}{\mathrm{d}z} \quad [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}]$')
axHisty.set_xlabel(r'$\frac{\mathrm{d}N}{\mathrm{d}S} \quad [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}]$')
#remove the lowest ticks from the histogram plots
#axHistx.set_yticks(axHistx.get_yticks()[:-1])
#axHisty.set_xticks(axHisty.get_xticks()[:-1])
#set minor ticks
axScatter.xaxis.set_minor_locator(MultipleLocator(0.5 / 5.))
axScatter.xaxis.set_minor_formatter(NullFormatter())
axScatter.yaxis.set_minor_locator(MultipleLocator(1. / 5.))
axScatter.yaxis.set_minor_formatter(NullFormatter())
#xhist
axHistx.xaxis.set_minor_locator(MultipleLocator(0.5 / 5.))
axHistx.xaxis.set_minor_formatter(NullFormatter())
#yhist
axHisty.yaxis.set_minor_locator(MultipleLocator(1. / 5.))
axHisty.yaxis.set_minor_formatter(NullFormatter())
plt.savefig(output)
def plot_luminosityfunction2(path,
database,
redshifts,
band,
out_folder,
solid_angle=10 * 160.,
ymin=10**3,
ymax=2 * 10**6,
xmin=0.5,
xmax=100,
nbins=15,
sigma=5.0,
H0=70.0,
WM=0.28,
zmax=6.0):
"""
:param solid_angle: area of the sky survey in arcmin**2
GOODS = 160, hence 10*160
:param sigma: sigma level of the errors to be plotted
:param nbins: number of bins (for simulated data)
"""
col = ['black', 'red', 'magenta', 'green', 'blue', 'brown']
#get data
query = '''select %s from FIR where %s > 7
and FIR.spire250_obs < 1e6''' % (band, band)
total = SamPy.db.sqlite.get_data_sqlite(path, database, query)
#make the figure
fig = P.figure()
ax = P.subplot(111)
#get the co-moving volume to the backend
comovingVol = cv.comovingVolume(solid_angle, 0, zmax, H0=H0, WM=WM)
#weight each galaxy
wghts = N.zeros(len(total)) + (1. / comovingVol)
#calculate the differential stellar mass function
#with log binning
b, n, nu = df.diff_function_log_binning(total,
wgth=wghts,
mmax=xmax,
mmin=xmin,
nbins=nbins,
log=True)
#calculate the poisson error
mask = nu > 0
err = wghts[0] * N.sqrt(nu[mask]) * sigma
up = n[mask] + err
lw = n[mask] - err
lw[lw < ymin] = ymin
#plot the knots
# mtot = ax.errorbar(b[mask], n[mask], yerr = [err, err],
# color = 'k', label = 'Total',
# marker = 'None', ls = '-')
#redshift limited plots
for i, red in enumerate(redshifts):
query = '''select %s from FIR where %s > 7 and %s
and FIR.spire250_obs < 1e6''' % (band, band, red)
limited = SamPy.db.sqlite.get_data_sqlite(path, database, query)
print query, len(limited)
#modify redshift string
tmp = red.split()
#rtitle = r'$z = %.0f$' % N.mean(N.array([float(tmp[2]), float(tmp[6])]))
rtitle = r'$%s < z \leq %s$' % (tmp[2], tmp[6])
#get a comoving volume
comovingVol = cv.comovingVolume(solid_angle,
float(tmp[2]),
float(tmp[6]),
H0=H0,
WM=WM)
#weights
wghts = N.zeros(len(limited)) + (1. / comovingVol)
#differential function
bb, nn, nu = df.diff_function_log_binning(limited,
wgth=wghts,
mmax=xmax,
mmin=xmin,
nbins=nbins,
log=True)
#calculate the poisson error
mask = nu > 0
# err = wghts[0] * N.sqrt(nu[mask]) * sigma
# up = nn[mask] + err
# lw = nn[mask] - err
# lw[lw < ymin] = ymin
x = bb[mask]
y = nn[mask]
#to make sure that the plots go below the area plotted
x = N.append(x, N.max(x) * 1.01)
y = N.append(y, 1e-10)
ax.plot(x, y, color=col[i], marker='None', ls='-', label=rtitle)
#set scales
ax.set_yscale('log')
ax.set_xlim(xmin + 0.2, xmax)
ax.set_ylim(ymin, ymax)
ax.set_ylabel(r'$\phi \ [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}]$')
ax.set_xlabel(r'$\log_{10}(L_{%s} \ [L_{\odot}])$' %
re.search('\d\d\d', band).group())
P.legend(scatterpoints=1, fancybox=True, shadow=True)
#save figure
P.savefig(out_folder + 'luminosity_function2_%s.ps' % band)
P.close()
def scatterHistograms(xdata, ydata, xlabel, ylabel, output):
"""
This functions generates a scatter plot and projected histograms to both axes.
"""
#constants
xmin = 2.0
xmax = 4.0
ymin = -2.05
ymax = 2.5
solid_angle = 100 * 160.
H0 = 70.0
WM = 0.28
#xbins and ybins
xbins = np.linspace(xmin, xmax, 30)
ybins = np.linspace(ymin, ymax, 20)
dfx = xbins[1] - xbins[0]
dfy = ybins[1] - ybins[0]
#calculate volume
comovingVol = cv.comovingVolume(solid_angle, xmin, xmax, H0=H0, WM=WM)
#weight each galaxy
wghtsx = (np.zeros(len(xdata)) + (1. / comovingVol)) / dfx
wghtsy = (np.zeros(len(ydata)) + (1. / comovingVol)) / dfy
#no labels, null formatter
nullfmt = NullFormatter()
# definitions for the axes
left, width = 0.1, 0.65
bottom, height = 0.1, 0.65
bottom_h = left_h = left + width + 0.02
rect_scatter = [left, bottom, width, height]
rect_histx = [left, bottom_h, width, 0.2]
rect_histy = [left_h, bottom, 0.2, height]
#make a figure
fig = plt.figure(figsize=(12, 12))
axScatter = plt.axes(rect_scatter)
axHistx = plt.axes(rect_histx)
axHisty = plt.axes(rect_histy)
#no labels
axHistx.xaxis.set_major_formatter(nullfmt)
axHisty.yaxis.set_major_formatter(nullfmt)
#the scatter plot
axScatter.scatter(xdata, ydata, marker='o', s=0.5, c='k')
#KDE
x = M.AnaKDE([xdata, ydata])
x_vec, y_vec, zm, lvls, d0, d1 = x.contour(
np.linspace(xmin - 0.1, xmax + 0.1, 50),
np.linspace(ymin - 0.1, ymax + 0.1, 50),
return_data=True)
axScatter.contourf(x_vec,
y_vec,
zm,
levels=np.linspace(0.002, 0.92 * np.max(zm), 10),
cmap=cm.get_cmap('gist_yarg'),
alpha=0.8)
#draw a line to the detection limit, 5mJy (at 250)
axScatter.axhline(np.log10(5), color='green', ls='--', lw=1.8)
#scatter labels
axScatter.set_xlabel(xlabel)
axScatter.set_ylabel(ylabel)
#set scatter limits
axScatter.set_xlim(xmin, xmax)
axScatter.set_ylim(ymin, ymax)
#make x histogram
x1 = axHistx.hist(xdata,
bins=xbins,
weights=wghtsx,
log=True,
color='gray')
x2 = axHistx.hist(xdata[ydata > np.log10(5)],
bins=xbins,
weights=wghtsx[ydata > np.log10(5)],
log=True,
color='gray',
hatch='x')
#set legend of x histogram
plt.legend((x1[2][0], x2[2][0]),
('All Galaxies', r'$S_{160}> 5\ \mathrm{mJy}$'),
shadow=False,
fancybox=False,
bbox_to_anchor=(0.01, 1.34),
loc=2,
borderaxespad=0.)
#make y histogram
axHisty.hist(ydata,
bins=ybins,
orientation='horizontal',
weights=wghtsy,
log=True,
color='gray')
#set histogram limits
axHistx.set_xlim(axScatter.get_xlim())
axHistx.set_ylim(1e-7, 1e-1)
axHisty.set_ylim(axScatter.get_ylim())
axHisty.set_xlim(4e-9, 1e-2)
#set histogram labels
axHistx.set_ylabel(
r'$\frac{\mathrm{d}N}{\mathrm{d}z} \quad [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}]$'
)
axHisty.set_xlabel(
r'$\frac{\mathrm{d}N}{\mathrm{d}S} \quad [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}]$'
)
#remove the lowest ticks from the histogram plots
#axHistx.set_yticks(axHistx.get_yticks()[:-1])
#axHisty.set_xticks(axHisty.get_xticks()[:-1])
#set minor ticks
axScatter.xaxis.set_minor_locator(MultipleLocator(0.5 / 5.))
axScatter.xaxis.set_minor_formatter(NullFormatter())
axScatter.yaxis.set_minor_locator(MultipleLocator(1. / 5.))
axScatter.yaxis.set_minor_formatter(NullFormatter())
#xhist
axHistx.xaxis.set_minor_locator(MultipleLocator(0.5 / 5.))
axHistx.xaxis.set_minor_formatter(NullFormatter())
#yhist
axHisty.yaxis.set_minor_locator(MultipleLocator(1. / 5.))
axHisty.yaxis.set_minor_formatter(NullFormatter())
plt.savefig(output)
def plot_luminosityfunctionPaper(path, database, redshifts,
bands, out_folder,
solid_angle=100*160.,
ymin=5e-7, ymax=5*10**-2,
xmin=9.1, xmax=13.1,
H0=70.0, WM=0.28):
"""
:param solid_angle: area of the sky survey in arcmin**2
GOODS = 160, hence 100*160
"""
col = ['black', 'red', 'magenta', 'green', 'blue']
lin = [':', '--', '-', '-.', '-']
#make the figure
fig = P.figure()
fig.subplots_adjust(left=0.09, bottom=0.08,
right=0.93, top=0.95,
wspace=0.0, hspace=0.0)
ax1 = fig.add_subplot(221)
ax2 = fig.add_subplot(222)
ax3 = fig.add_subplot(223)
ax4 = fig.add_subplot(224)
for b in bands:
if '100' in b: nb = 19
if '160' in b: nb = 14
if '250' in b: nb = 14
if '350' in b: nb = 11
print '\nPlotting ', b
#redshift limited plots
for i, red in enumerate(redshifts):
query = '''select %s from FIR where %s > 7.7 and %s
and FIR.spire250_obs < 1e6''' % (b, b, red)
limited = SamPy.db.sqlite.get_data_sqlite(path, database, query)
print query, len(limited)
#modify redshift string
tmp = red.split()
#rtitle = r'$z = %.0f$' % N.mean(N.array([float(tmp[2]), float(tmp[6])]))
rtitle = r'$%s \leq z < %s$' % (tmp[2], tmp[6])
#get a comoving volume
comovingVol = cv.comovingVolume(solid_angle,
float(tmp[2]),
float(tmp[6]),
H0=H0,
WM=WM)
#weights
wghts = N.zeros(len(limited)) + (1. / comovingVol)
#differential function
bb, nn, nu = df.diff_function_log_binning(limited,
wgth=wghts,
mmax=xmax,
mmin=xmin,
nbins=nb,
log=True)
mask = nu > 0
x = bb[mask]
y = nn[mask]
#to make sure that the plots go below the area plotted
x = N.append(x, N.max(x) * 1.01)
y = N.append(y, 1e-10)
if '100' in b:
ax1.plot(x, N.log10(y), color=col[i], marker='None', ls=lin[i], label=rtitle)
if '160' in b:
ax2.plot(x, N.log10(y), color=col[i], marker='None', ls=lin[i], label=rtitle)
if '250' in b:
ax3.plot(x, N.log10(y), color=col[i], marker='None', ls=lin[i], label=rtitle)
if '350' in b:
ax4.plot(x, N.log10(y), color=col[i], marker='None', ls=lin[i], label=rtitle)
#plot observational constrains
mic100 = lf.Herschel100Lapi()
mic250 = lf.Herschel250Lapi()
#x values
x100 = N.log10(mic100['Lsun'])
x250 = N.log10(mic250['Lsun'])
#mask out missing values
msk100z15 = mic100['z1.5'][:, 0] > -6.5
msk100z22 = mic100['z2.2'][:, 0] > -6.5
msk100z32 = mic100['z3.2'][:, 0] > -6.5
msk250z15 = mic250['z1.4'][:, 0] > -6.5
msk250z22 = mic250['z2.2'][:, 0] > -6.5
msk250z32 = mic250['z3.2'][:, 0] > -6.5
#PACS100 plots
ax1.errorbar(x100[msk100z15], mic100['z1.5'][:, 0][msk100z15],
yerr=[-mic100['z1.5'][:, 2][msk100z15], mic100['z1.5'][:, 1][msk100z15]],
marker='s', ms=4, ls='None', mfc='r', mec='r', c='r')
ax1.errorbar(x100[msk100z22], mic100['z2.2'][:, 0][msk100z22],
yerr=[-mic100['z2.2'][:, 2][msk100z22], mic100['z2.2'][:, 1][msk100z22]],
marker='o', ms=4, ls='None', mfc='m', mec='m', c='m')
ax1.errorbar(x100[msk100z32], mic100['z3.2'][:, 0][msk100z32],
yerr=[-mic100['z3.2'][:, 2][msk100z32], mic100['z3.2'][:, 1][msk100z32]],
marker='d', ms=4, ls='None', mfc='g', mec='g', c='g')
# ax1.scatter(x100[msk100z15], 10**mic100['z1.5'][:,0][msk100z15],
# marker='s', s=10,c='k')
# ax1.scatter(x100[msk100z22], 10**mic100['z2.2'][:,0][msk100z22],
# marker='o', s=10, c='r')
# ax1.scatter(x100[msk100z32], 10**mic100['z3.2'][:,0][msk100z32],
# marker='d', s=10, c='m')
#SPIRE250 plots
ax3.errorbar(x250[msk250z15], mic250['z1.4'][:, 0][msk250z15],
yerr=[-mic250['z1.4'][:, 2][msk250z15], mic250['z1.4'][:, 1][msk250z15]],
marker='s', ms=4, ls='None', mfc='r', mec='r', c='r', label=r'$1.2 \leq z < 1.6$')
ax3.errorbar(x250[msk250z22], mic250['z2.2'][:, 0][msk250z22],
yerr=[-mic250['z2.2'][:, 2][msk250z22], mic250['z2.2'][:, 1][msk250z22]],
marker='o', ms=4, ls='None', mfc='m', mec='m', c='m', label=r'$2.0 \leq z < 2.4$')
ax3.errorbar(x250[msk250z32], mic250['z3.2'][:, 0][msk250z32],
yerr=[-mic250['z3.2'][:, 2][msk250z32], mic250['z3.2'][:, 1][msk250z32]],
marker='d', ms=4, ls='None', mfc='g', mec='g', c='g', label=r'$2.4 \leq z < 4.0$')
# ax3.scatter(x250[msk250z15], 10**mic250['z1.4'][:,0][msk250z15],
# marker='s', s=10,c='k')
# ax3.scatter(x250[msk250z22], 10**mic250['z2.2'][:,0][msk250z22],
# marker='o', s=10, c='r')
# ax3.scatter(x250[msk250z32], 10**mic250['z3.2'][:,0][msk250z32],
# marker='d', s=10, c='m')
#labels
ax4.errorbar([5,], [-10,], yerr=[0.1,],
marker='s', ms=4, ls='None', mfc='r', mec='r', c='r', label=r'$1.2 \leq z < 1.6$')
ax4.errorbar([5,], [-10,], yerr=[0.1,],
marker='o', ms=4, ls='None', mfc='m', mec='m', c='m', label=r'$2.0 \leq z < 2.4$')
ax4.errorbar([5,], [-10,], yerr=[0.1,],
marker='d', ms=4, ls='None', mfc='g', mec='g', c='g', label=r'$2.4 \leq z < 4.0$')
#set scales
# ax1.set_yscale('log')
# ax2.set_yscale('log')
# ax3.set_yscale('log')
# ax4.set_yscale('log')
#ylabel = r'$\phi \ [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}]$'
ylabel = r'$\log_{10} \left ( \phi \ [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}] \right )$'
xlabel = r'$\log_{10}(L \ [L_{\odot}])$'
#labels
ax3.set_xlabel(xlabel)
ax4.set_xlabel(xlabel)
ax1.set_ylabel(ylabel)
ax3.set_ylabel(ylabel)
ax2.set_yticklabels([])
ax4.set_yticklabels([])
ax1.set_xticklabels([])
ax2.set_xticklabels([])
#limits
ymin = N.log10(ymin)
ymax = N.log10(ymax)
ax1.set_ylim(ymin, ymax)
ax1.set_xlim(xmin + 0.2, xmax)
ax2.set_ylim(ymin, ymax)
ax2.set_xlim(xmin + 0.2, xmax)
ax3.set_ylim(ymin, ymax)
ax3.set_xlim(xmin + 0.2, xmax)
ax4.set_ylim(ymin, ymax)
ax4.set_xlim(xmin + 0.2, xmax)
#add some annotations
P.text(0.5, 0.94, 'a) PACS 100',
horizontalalignment='center',
verticalalignment='center',
transform=ax1.transAxes)
P.text(0.5, 0.94, 'b) PACS 160',
horizontalalignment='center',
verticalalignment='center',
transform=ax2.transAxes)
P.text(0.5, 0.94, 'c) SPIRE 250',
horizontalalignment='center',
verticalalignment='center',
transform=ax3.transAxes)
P.text(0.5, 0.94, 'd) SPIRE 350',
horizontalalignment='center',
verticalalignment='center',
transform=ax4.transAxes)
#legend
ax4.legend(scatterpoints=1, fancybox=True, shadow=True,
loc='center right')
#save figure
P.savefig(out_folder + 'luminosity_functionPaper.ps')
P.close()
def plot_luminosityfunction(path, database, redshifts,
band, out_folder,
solid_angle=10 * 160.,
ymin=10 ** 3, ymax=2 * 10 ** 6,
xmin=0.5, xmax=100,
nbins=15, sigma=5.0,
H0=70.0, WM=0.28,
zmax=6.0):
"""
:param solid_angle: area of the sky survey in arcmin**2
GOODS = 160, hence 10*160
:param sigma: sigma level of the errors to be plotted
:param nbins: number of bins (for simulated data)
"""
#subplot numbers
columns = 3
rows = 3
#get data
query = """select %s from FIR where %s > 7
and FIR.spire250_obs < 1e6""" % (band, band)
total = SamPy.db.sqlite.get_data_sqlite(path, database, query)
#make the figure
fig = P.figure()
P.subplots_adjust(wspace=0.0, hspace=0.0)
ax = P.subplot(rows, columns, 1)
#get the co-moving volume to the backend
comovingVol = cv.comovingVolume(solid_angle, 0, zmax,
H0=H0, WM=WM)
#weight each galaxy
wghts = N.zeros(len(total)) + (1. / comovingVol)
#calculate the differential stellar mass function
#with log binning
b, n, nu = df.diff_function_log_binning(total,
wgth=wghts,
mmax=xmax,
mmin=xmin,
nbins=nbins,
log=True)
#calculate the poisson error
mask = nu > 0
err = wghts[0] * N.sqrt(nu[mask]) * sigma
up = n[mask] + err
lw = n[mask] - err
lw[lw < ymin] = ymin
#plot the sigma area
stot = ax.fill_between(b[mask], up, lw, color='#728FCE')
#plot the knots
mtot = ax.scatter(b[mask], n[mask], marker='o', s=3, color='k')
#add annotation
ax.annotate('Total', (0.5, 0.87), xycoords='axes fraction',
ha='center')
#set scale
ax.set_yscale('log')
ax.set_xlim(xmin, xmax)
ax.set_ylim(ymin, ymax)
ax.set_xticklabels([])
ax.set_ylabel(r'$\phi \ [Mpc^{-3} \ dex^{-1}]$')
ptot = P.Rectangle((0, 0), 1, 1, fc='#728FCE')
sline = '%i$\sigma$ errors' % sigma
P.legend((mtot, ptot), ('All Galaxies', sline), loc='lower left',
scatterpoints=1, fancybox=True, shadow=True)
#redshift limited plots
for i, red in enumerate(redshifts):
query = '''select %s from FIR where %s > 7 and %s
and FIR.spire250_obs < 1e6''' % (band, band, red)
limited = SamPy.db.sqlite.get_data_sqlite(path, database, query)
print query, len(limited)
#modify redshift string
tmp = red.split()
#rtitle = r'$z = %.0f$' % N.mean(N.array([float(tmp[2]), float(tmp[6])]))
rtitle = r'$%s < z \leq %s$' % (tmp[2], tmp[6])
#get a comoving volume
comovingVol = cv.comovingVolume(solid_angle,
float(tmp[2]),
float(tmp[6]),
H0=H0,
WM=WM)
#weights
wghts = N.zeros(len(limited)) + (1. / comovingVol)
#differential function
bb, nn, nu = df.diff_function_log_binning(limited,
wgth=wghts,
mmax=xmax,
mmin=xmin,
nbins=nbins,
log=True)
#make a subplot
axs = P.subplot(rows, columns, i + 2)
#calculate the poisson error
mask = nu > 0
err = wghts[0] * N.sqrt(nu[mask]) * sigma
up = nn[mask] + err
lw = nn[mask] - err
lw[lw < ymin] = ymin
#plot the sigma area
axs.fill_between(bb[mask], up, lw, color='#728FCE')
#plot the knots
axs.scatter(bb[mask], nn[mask], marker='o',
s=3, color='k')
#add annotation
axs.annotate(rtitle, (0.5, 0.87),
xycoords='axes fraction',
ha='center')
#set scales
axs.set_yscale('log')
axs.set_xlim(xmin, xmax)
axs.set_ylim(ymin, ymax)
#remove unnecessary ticks and add units
if i == 0 or i == 1 or i == 3 or i == 4:
axs.set_yticklabels([])
if i == 2 or i == 3 or i == 4:
btmp = re.search('\d\d\d', band).group()
axs.set_xlabel(r'$\log_{10} (L_{%s} \ [L_{\odot}])$' % btmp)
#axs.set_xticks(axs.get_xticks()[1:])
else:
axs.set_xticklabels([])
if i == 2:
axs.set_ylabel(r'$\phi \ [Mpc^{-3} \ dex^{-1}]$')
#axs.set_xticks(axs.get_xticks()[:-1])
#save figure
P.savefig(out_folder + 'luminosity_function_%s.ps' % band)
P.close()
def plot_luminosityfunctionKDE(path, database, redshifts,
band, out_folder,
solid_angle=10 * 160.,
ymin=10 ** 3, ymax=2 * 10 ** 6,
xmin=0.5, xmax=100,
nbins=15,
H0=70.0, WM=0.28,
zmax=6.0):
"""
:param solid_angle: area of the sky survey in arcmin**2
GOODS = 160, hence 10*160
:param sigma: sigma level of the errors to be plotted
:param nbins: number of bins (for simulated data)
"""
col = ['black', 'red', 'magenta', 'green', 'blue', 'brown']
#get data
query = '''select %s from FIR where %s > 6
and FIR.spire250_obs < 1e6''' % (band, band)
total = SamPy.db.sqlite.get_data_sqlite(path, database, query)[:, 0]
print len(total)
#get the co-moving volume to the backend
comovingVol = cv.comovingVolume(solid_angle,
0, zmax,
H0=H0, WM=WM)
#normalization
normalization = float(len(total)) / comovingVol * (nbins * 7 * 2)
#KDE
mu = SS.gaussian_kde(total)
#in which points to evaluate
x = N.linspace(N.min(total), N.max(total), nbins * 7)
#make the figure
fig = P.figure()
ax = P.subplot(111)
#plot
ax.plot(x, mu.evaluate(x) / normalization, color='gray', ls='--')
#redshift limited plots
for i, red in enumerate(redshifts):
query = '''select %s from FIR where %s > 6 and %s
and FIR.spire250_obs < 1e6''' % (band, band, red)
limited = SamPy.db.sqlite.get_data_sqlite(path, database, query)[:, 0]
print query, len(limited)
#modify redshift string
tmp = red.split()
rtitle = r'$%s < z \leq %s$' % (tmp[2], tmp[6])
#get a comoving volume
comovingVol = cv.comovingVolume(solid_angle,
float(tmp[2]),
float(tmp[6]),
H0=H0,
WM=WM)
#normalization
normalization = float(len(limited)) / comovingVol * (nbins * 7 * 2)
#KDE
mu = SS.gaussian_kde(limited)
#in which points to evaluate
x = N.linspace(N.min(limited), N.max(limited), nbins * 7)
ax.plot(x, mu.evaluate(x) / normalization, color=col[i],
marker='None', ls='-', label=rtitle)
#set scales
ax.set_yscale('log')
ax.set_xlim(xmin, xmax)
ax.set_ylim(ymin, ymax)
ax.set_ylabel(r'$\phi \ [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}]$')
ax.set_xlabel(r'$\log_{10}(L_{%s} \ [L_{\odot}])$' % re.search('\d\d\d', band).group())
P.legend(scatterpoints=1, fancybox=True, shadow=True)
#save figure
P.savefig(out_folder + 'luminosity_functionKDE_%s.ps' % band)
P.close()
def plot_luminosityfunction2(path, database, redshifts,
band, out_folder,
solid_angle=10 * 160.,
ymin=10 ** 3, ymax=2 * 10 ** 6,
xmin=0.5, xmax=100,
nbins=15, sigma=5.0,
H0=70.0, WM=0.28,
zmax=6.0):
"""
:param solid_angle: area of the sky survey in arcmin**2
GOODS = 160, hence 10*160
:param sigma: sigma level of the errors to be plotted
:param nbins: number of bins (for simulated data)
"""
col = ['black', 'red', 'magenta', 'green', 'blue', 'brown']
#get data
query = '''select %s from FIR where %s > 7
and FIR.spire250_obs < 1e6''' % (band, band)
total = SamPy.db.sqlite.get_data_sqlite(path, database, query)
#make the figure
fig = P.figure()
ax = P.subplot(111)
#get the co-moving volume to the backend
comovingVol = cv.comovingVolume(solid_angle, 0, zmax,
H0=H0, WM=WM)
#weight each galaxy
wghts = N.zeros(len(total)) + (1. / comovingVol)
#calculate the differential stellar mass function
#with log binning
b, n, nu = df.diff_function_log_binning(total,
wgth=wghts,
mmax=xmax,
mmin=xmin,
nbins=nbins,
log=True)
#calculate the poisson error
mask = nu > 0
err = wghts[0] * N.sqrt(nu[mask]) * sigma
up = n[mask] + err
lw = n[mask] - err
lw[lw < ymin] = ymin
#plot the knots
# mtot = ax.errorbar(b[mask], n[mask], yerr = [err, err],
# color = 'k', label = 'Total',
# marker = 'None', ls = '-')
#redshift limited plots
for i, red in enumerate(redshifts):
query = '''select %s from FIR where %s > 7 and %s
and FIR.spire250_obs < 1e6''' % (band, band, red)
limited = SamPy.db.sqlite.get_data_sqlite(path, database, query)
print query, len(limited)
#modify redshift string
tmp = red.split()
#rtitle = r'$z = %.0f$' % N.mean(N.array([float(tmp[2]), float(tmp[6])]))
rtitle = r'$%s < z \leq %s$' % (tmp[2], tmp[6])
#get a comoving volume
comovingVol = cv.comovingVolume(solid_angle,
float(tmp[2]),
float(tmp[6]),
H0=H0,
WM=WM)
#weights
wghts = N.zeros(len(limited)) + (1. / comovingVol)
#differential function
bb, nn, nu = df.diff_function_log_binning(limited,
wgth=wghts,
mmax=xmax,
mmin=xmin,
nbins=nbins,
log=True)
#calculate the poisson error
mask = nu > 0
# err = wghts[0] * N.sqrt(nu[mask]) * sigma
# up = nn[mask] + err
# lw = nn[mask] - err
# lw[lw < ymin] = ymin
x = bb[mask]
y = nn[mask]
#to make sure that the plots go below the area plotted
x = N.append(x, N.max(x) * 1.01)
y = N.append(y, 1e-10)
ax.plot(x, y, color=col[i], marker='None', ls='-', label=rtitle)
#set scales
ax.set_yscale('log')
ax.set_xlim(xmin + 0.2, xmax)
ax.set_ylim(ymin, ymax)
ax.set_ylabel(r'$\phi \ [\mathrm{Mpc}^{-3} \ \mathrm{dex}^{-1}]$')
ax.set_xlabel(r'$\log_{10}(L_{%s} \ [L_{\odot}])$' % re.search('\d\d\d', band).group())
P.legend(scatterpoints=1, fancybox=True, shadow=True)
#save figure
P.savefig(out_folder + 'luminosity_function2_%s.ps' % band)
P.close()