def files_of_mass_weights(log_age, metal, weights, probs, filename=""):
best_fit = np.argmax(probs)
weights = weights
fits_arr = [best_fit]
for i, fits in enumerate(fits_arr):
if i == 0:
ii = (0, 0)
inc_weights = statistics.renormalise_weights(weights[fits_arr],
probs[fits_arr])
inc_weights = inc_weights / np.sum(inc_weights)
prev_weight = np.zeros(len(weights[0]))
tot_weight = np.sum(inc_weights, 0)
for i in range(len(log_age)):
file = open('./' + str(filename) + '_age.txt', 'a')
file.write('%12.6f\t' % log_age[i])
file.write('\n')
file.close()
for i in range(len(metal)):
file = open('./' + str(filename) + '_metal.txt', 'a')
file.write('%12.6f\t' % metal[i])
file.write('\n')
file.close()
for i in range(len(probs)):
file = open('./' + str(filename) + '_mass_probs.txt', 'a')
file.write('%12.6f\t' % probs[i])
file.write('\n')
file.close()
for i in range(len(metal)):
file = open('./' + str(filename) + '_mass_weights.txt', 'a')
file.write('%12.6f\t' % tot_weight[i])
file.write('\n')
file.close()
def files_of_mass_weights(log_age, metal, weights, probs, filename=""):
best_fit = np.argmax(probs)
weights=weights
fits_arr = [best_fit]
for i,fits in enumerate(fits_arr):
if i == 0:
ii = (0,0)
inc_weights = statistics.renormalise_weights(weights[fits_arr],probs[fits_arr])
inc_weights = inc_weights / np.sum(inc_weights)
prev_weight = np.zeros(len(weights[0]))
tot_weight = np.sum(inc_weights,0)
for i in range(len(log_age)):
file=open('./'+str(filename)+'_age.txt', 'a')
file.write('%12.6f\t' % log_age[i])
file.write('\n')
file.close()
for i in range(len(metal)):
file=open('./'+str(filename)+'_metal.txt', 'a')
file.write('%12.6f\t' % metal[i])
file.write('\n')
file.close()
for i in range(len(probs)):
file=open('./'+str(filename)+'_mass_probs.txt', 'a')
file.write('%12.6f\t' % probs[i])
file.write('\n')
file.close()
for i in range(len(metal)):
file=open('./'+str(filename)+'_mass_weights.txt', 'a')
file.write('%12.6f\t' % tot_weight[i])
file.write('\n')
file.close()
def averages_and_errors(probs, prop, sampling):
def max_pdf(probs, prop, sampling):
lower_limit = np.min(prop)
upper_limit = np.max(prop)
if upper_limit == lower_limit:
return np.asarray(prop), np.ones(len(probs)) / np.size(probs)
property_pdf_int = np.arange(
lower_limit, upper_limit * 1.001, (upper_limit - lower_limit) /
sampling) + (upper_limit - lower_limit) * 0.00000001
prob_pdf = np.zeros(len(property_pdf_int))
def bisect_array(array):
bisected_array = np.zeros(len(array) - 1)
for ai in range(len(bisected_array)):
bisected_array[ai] = (array[ai] + array[ai + 1]) / 2.0
return bisected_array
for p in range(len(property_pdf_int) - 1):
match_prop = np.where((prop <= property_pdf_int[p + 1])
& (prop > property_pdf_int[p]))
if np.size(match_prop) == 0:
continue
else:
prob_pdf[p] = np.max(probs[match_prop])
property_pdf = bisect_array(property_pdf_int)
return property_pdf, prob_pdf[:-1] / np.sum(prob_pdf)
print ' we are now in the averages section.'
best_fit = np.argmax(probs)
one_sig_fits = np.where(probs > (1 - 0.6827))
weights = light_weights
fits_arr = [best_fit]
for i, fits in enumerate(fits_arr):
if i == 0:
ii = (0, 0)
inc_weights = statistics.renormalise_weights(
weights[fits_arr], probs[fits_arr])
inc_weights = inc_weights / np.sum(inc_weights)
prev_weight = np.zeros(len(weights[0]))
tot_weight = np.sum(inc_weights, 0)
print sampling
print '-'
xdf, y = max_pdf(probs, prop, sampling)
cdf = np.zeros(np.shape(y))
cdf_probspace = np.zeros(np.shape(y))
for m in range(len(y)):
cdf[m] = np.sum(y[:m])
cdf = cdf / np.max(cdf)
#fig,ax=plt.subplots()
#ax.tick_params(labelsize=20)
#ax.plot(prop+9.0,probs/np.max(probs),'k.',label='Individual fits')
#ax.plot(xdf+9.0,y/np.max(y),'r--',linewidth=2.0,label='Likelihood distribution')
#xlim_in = (np.min(xdf[np.where(y/np.max(y)>0.01)])+9.0,np.max(xdf[np.where(y/np.max(y)>0.01)])+9.0)
#plt.plot(xdf,cdf,'b-')
#ax.set_ylabel("Likelihood",fontsize=22)
#ax.set_xlabel("Age / log(yrs)",fontsize=22)
#ax.legend(loc='upper left', shadow=True,prop={'size':20})
#out_plot_string = './plots/likelihood_'+str(np.mean(prop))+'.eps'
#plt.savefig(out_plot_string,format='eps',transparent=True,bbox_inches='tight')
#plt.close()
area_probspace = y * (xdf[1] - xdf[0])
area_probspace = area_probspace / np.sum(area_probspace)
indx_probspace = np.argsort(area_probspace)[::-1]
desc_probspace = np.sort(area_probspace)[::-1]
cdf_probspace = np.zeros(np.shape(desc_probspace))
for m in range(len(desc_probspace)):
cdf_probspace[m] = np.sum(desc_probspace[:m])
def find_closest(A, target):
#A must be sorted
idx = A.searchsorted(target)
idx = np.clip(idx, 1, len(A) - 1)
left = A[idx - 1]
right = A[idx]
idx -= target - left < right - target
return idx
# Median, + / - 1 sig, + / - 2 sig, + / - 3 sig
#av_sigs = [0.5,0.6827/2.0+0.5,0.5-0.6827/2.0,0.95/2.0+0.5,0.5-0.95/2.0,0.997/2.0+0.5,0.5-0.997/2.0]
av_sigs = [0.6827, 0.9545, 0.9973]
# Sorts results by likelihood and calculates confidence intervals on sorted space
index_close = find_closest(cdf_probspace, av_sigs)
best_fit = xdf[indx_probspace[0]]
upper_onesig, lower_onesig = np.max(
xdf[indx_probspace[:index_close[0]]]), np.min(
xdf[indx_probspace[:index_close[0]]])
upper_twosig, lower_twosig = np.max(
xdf[indx_probspace[:index_close[1]]]), np.min(
xdf[indx_probspace[:index_close[1]]])
upper_thrsig, lower_thrsig = np.max(
xdf[indx_probspace[:index_close[2]]]), np.min(
xdf[indx_probspace[:index_close[2]]])
# Takes whole pdf and computes median, confidence intervals
#index_close = find_closest(cdf, av_sigs)
if np.size(xdf) == 0:
raise Exception(
'No solutions found??? FIREFLY error (see statistics.py)')
#return xdf[index_close]
return [
best_fit, upper_onesig, lower_onesig, upper_twosig, lower_twosig,
upper_thrsig, lower_thrsig
]
def averages_and_errors(probs,prop,sampling):
def max_pdf(probs,prop,sampling):
lower_limit = np.min(prop)
upper_limit = np.max(prop)
if upper_limit==lower_limit:
return np.asarray(prop),np.ones(len(probs))/np.size(probs)
property_pdf_int= np.arange(lower_limit,upper_limit*1.001,(upper_limit-lower_limit)/sampling)+(upper_limit-lower_limit)*0.00000001
prob_pdf = np.zeros(len(property_pdf_int))
def bisect_array(array):
bisected_array = np.zeros(len(array)-1)
for ai in range(len(bisected_array)):
bisected_array[ai] = (array[ai]+array[ai+1])/2.0
return bisected_array
for p in range(len(property_pdf_int)-1):
match_prop = np.where( (prop <= property_pdf_int[p+1]) & (prop > property_pdf_int[p]) )
if np.size(match_prop) == 0:
continue
else:
prob_pdf[p] = np.max( probs[match_prop] )
property_pdf = bisect_array(property_pdf_int)
return property_pdf,prob_pdf[:-1]/np.sum(prob_pdf)
print' we are now in the averages section.'
best_fit = np.argmax(probs)
one_sig_fits = np.where(probs>(1-0.6827))
weights=light_weights
fits_arr = [best_fit]
for i,fits in enumerate(fits_arr):
if i == 0:
ii = (0,0)
inc_weights = statistics.renormalise_weights(weights[fits_arr],probs[fits_arr])
inc_weights = inc_weights / np.sum(inc_weights)
prev_weight = np.zeros(len(weights[0]))
tot_weight = np.sum(inc_weights,0)
print sampling
print '-'
xdf,y = max_pdf(probs,prop,sampling)
cdf = np.zeros(np.shape(y))
cdf_probspace = np.zeros(np.shape(y))
for m in range(len(y)):
cdf[m] = np.sum(y[:m])
cdf = cdf / np.max(cdf)
#fig,ax=plt.subplots()
#ax.tick_params(labelsize=20)
#ax.plot(prop+9.0,probs/np.max(probs),'k.',label='Individual fits')
#ax.plot(xdf+9.0,y/np.max(y),'r--',linewidth=2.0,label='Likelihood distribution')
#xlim_in = (np.min(xdf[np.where(y/np.max(y)>0.01)])+9.0,np.max(xdf[np.where(y/np.max(y)>0.01)])+9.0)
#plt.plot(xdf,cdf,'b-')
#ax.set_ylabel("Likelihood",fontsize=22)
#ax.set_xlabel("Age / log(yrs)",fontsize=22)
#ax.legend(loc='upper left', shadow=True,prop={'size':20})
#out_plot_string = './plots/likelihood_'+str(np.mean(prop))+'.eps'
#plt.savefig(out_plot_string,format='eps',transparent=True,bbox_inches='tight')
#plt.close()
area_probspace = y*(xdf[1]-xdf[0])
area_probspace = area_probspace/np.sum(area_probspace)
indx_probspace = np.argsort(area_probspace)[::-1]
desc_probspace = np.sort(area_probspace)[::-1]
cdf_probspace = np.zeros(np.shape(desc_probspace))
for m in range(len(desc_probspace)):
cdf_probspace[m] = np.sum(desc_probspace[:m])
def find_closest(A, target):
#A must be sorted
idx = A.searchsorted(target)
idx = np.clip(idx, 1, len(A)-1)
left = A[idx-1]
right = A[idx]
idx -= target - left < right - target
return idx
# Median, + / - 1 sig, + / - 2 sig, + / - 3 sig
#av_sigs = [0.5,0.6827/2.0+0.5,0.5-0.6827/2.0,0.95/2.0+0.5,0.5-0.95/2.0,0.997/2.0+0.5,0.5-0.997/2.0]
av_sigs = [0.6827,0.9545,0.9973]
# Sorts results by likelihood and calculates confidence intervals on sorted space
index_close = find_closest(cdf_probspace, av_sigs)
best_fit = xdf[indx_probspace[0]]
upper_onesig,lower_onesig = np.max(xdf[indx_probspace[:index_close[0]]]),np.min(xdf[indx_probspace[:index_close[0]]])
upper_twosig,lower_twosig = np.max(xdf[indx_probspace[:index_close[1]]]),np.min(xdf[indx_probspace[:index_close[1]]])
upper_thrsig,lower_thrsig = np.max(xdf[indx_probspace[:index_close[2]]]),np.min(xdf[indx_probspace[:index_close[2]]])
# Takes whole pdf and computes median, confidence intervals
#index_close = find_closest(cdf, av_sigs)
if np.size(xdf) == 0:
raise Exception('No solutions found??? FIREFLY error (see statistics.py)')
#return xdf[index_close]
return [best_fit,upper_onesig,lower_onesig,upper_twosig,lower_twosig,upper_thrsig,lower_thrsig]
def plot_sfh(age,metal,weights,probs):
# Plot histograms of the SFH of the best fit, 1 sig, (2 sig?) solutions.
# (Maybe with CDF)
# INPUT: ages, metallicities, weights, probs
age = np.log10(age)+9.0
def diff_array(array):
bisected_array = np.zeros(len(array)-1)
for ai in range(len(bisected_array)):
bisected_array[ai] = (array[ai+1]-array[ai])
return bisected_array
min_width = np.min(np.absolute((diff_array(age))))
best_fit = np.argmax(probs)
one_sig_fits = np.where(probs>(1-0.6827))
two_sig_fits = np.where(probs>(1-0.9545))
fits_arr = [[best_fit],one_sig_fits,two_sig_fits]
print fits_arr
sort_ages = np.sort(age)
sort_age_index = np.argsort(age)
sort_unique_ages = np.sort(np.unique(age))
sort_metals = np.sort(metal)
sort_unique_metals = np.sort(np.unique(metal))
color_metals = ['violet','blue','MediumAquaMarine','green','yellow','orange','darkred']
name_metals = [str(x) for x in sort_unique_metals]
f, ax_arr = plt.subplots(3, sharex=True, sharey=False)
ax_cdf = [ax_arr[0].twinx(),ax_arr[1].twinx(),ax_arr[2].twinx()]
ax_arr[0].plot([-99,-99],[0,0],'w.',label='[Z/H] = ')
for nm in range(len(name_metals)):
if name_metals[nm]=='-2.302':
ax_arr[0].bar(-99,0,facecolor=color_metals[nm],label='-2.3 (RHB)')
elif name_metals[nm]=='-2.301':
ax_arr[0].bar(-99,0,facecolor=color_metals[nm],label='-2.3 (BHB)')
elif name_metals[nm]=='-1.302':
ax_arr[0].bar(-99,0,facecolor=color_metals[nm],label='-1.3 (RHB)')
elif name_metals[nm]=='-1.301':
ax_arr[0].bar(-99,0,facecolor=color_metals[nm],label='-1.3 (BHB)')
else:
ax_arr[0].bar(-99,0,facecolor=color_metals[nm],label=name_metals[nm])
for i,fits in enumerate(fits_arr):
#inc_weights = weights[fits]
import statistics
inc_weights = statistics.renormalise_weights(weights[fits],probs[fits])
inc_weights = inc_weights / np.sum(inc_weights)
prev_weight = np.zeros(len(weights[0]))
tot_weight = np.sum(inc_weights,0)
print tot_weight
for mi,m in enumerate(sort_unique_metals):
for w in range(len(weights[0])):
if metal[w] == m:
ax_arr[i].bar(age[w],tot_weight[w],width=min_width,facecolor=color_metals[mi],bottom=prev_weight[w],align='center')
prev_weight[w] = tot_weight[w]+prev_weight[w]
# Make a CDF:
pdf_bar = np.zeros(len(sort_unique_ages))
cdf_bar = np.zeros(len(sort_unique_ages))
for a in range(len(sort_unique_ages)):
pdf_bar[a] = np.sum(prev_weight[np.where(age==sort_unique_ages[a])])
cdf_bar[a] = np.sum(pdf_bar[:a])
ax_cdf[i].plot([np.min(sort_unique_ages)-99,np.min(sort_unique_ages)]+sort_unique_ages.tolist()+[np.max(sort_unique_ages),np.max(sort_unique_ages)+99],\
[0,0]+cdf_bar.tolist()+[1.0,1.0],'--',linewidth=4.0,color='springgreen')
# Fine-tune figure; make subplots close to each other and hide x ticks for
# all but bottom plot.c
ax_arr[1].set_ylabel("SSP Weight")
ax_arr[-1].set_xlabel("Age / log(yr)")
ax_cdf[1].set_ylabel("CDF of SSP weight")
handles, labels = ax_arr[0].get_legend_handles_labels()
ax_arr[0].legend(handles,labels,shadow=True,ncol=int(round(len(labels)/2)),bbox_to_anchor=(0.01, 0.95, 0.98, .105), loc=3,\
mode="expand", borderaxespad=0.,handlelength=0.6,handleheight=1.2,handletextpad=0.4,\
framealpha=1.0)
ax_arr[-1].set_xlim(np.min(age[np.where(prev_weight>0.001)])-min_width,np.max(age[np.where(prev_weight>0.001)])+min_width)
f.subplots_adjust(hspace=0)
plt.setp([a.get_xticklabels() for a in ax_arr[:-1]], visible=False)
plt.show()
def plot_sfh_contours(age,metal,weights,probs,title=""):
# Plot FIREFLY'S likelihood-weighted SSP contribution contours
# INPUT: ages, metallicities, weights, probs
age = np.log10(age)+9.0
metal = np.asarray(metal)
def diff_array(array):
bisected_array = np.zeros(len(array)-1)
for ai in range(len(bisected_array)):
bisected_array[ai] = (array[ai+1]-array[ai])
return bisected_array
min_width = np.min(np.absolute((diff_array(age))))
best_fit = np.argmax(probs)
one_sig_fits = np.where(probs>(1-0.6827))
two_sig_fits = np.where(probs>(1-0.9545))
fits_arr = [[best_fit],one_sig_fits,two_sig_fits]
label_arr = ["Best fit","68% level","95% level"]
col_cum = ['black','green','cyan']
sort_ages = np.sort(age)
sort_age_index = np.argsort(age)
sort_unique_ages = np.sort(np.unique(age))
sort_metals = np.sort(metal)
sort_unique_metals = np.sort(np.unique(metal))
name_metals = [str(x) for x in sort_unique_metals]
f, ax_arr = plt.subplots(2, 2, sharex=True, sharey=True)
# ax_arr = ( (1,2),(3,4) )
def truncate_colormap(cmap, minval=0.0, maxval=1.0, n=100):
new_cmap = colors.LinearSegmentedColormap.from_list(
'trunc({n},{a:.2f},{b:.2f})'.format(n=cmap.name, a=minval, b=maxval),
cmap(np.linspace(minval, maxval, n)))
return new_cmap
cmap=truncate_colormap(get_cmap('hot_r'),minval=0.15,maxval=1.0,n=1000)
cmap.set_under('w')
bounds_interp_age = (9.0,10.19)
bounds_interp_metal = (-2.5,0.90)
ax_arr[0,1]= plt.subplot(222)
ax_arr[0,1].set_xlim(bounds_interp_age[0],bounds_interp_age[1])
#ax_arr[0,1].set_xlabel("Age / log(yr)",fontsize=20)
for i,fits in enumerate(fits_arr):
#ii = (int(i/2),i%2)
if i == 0:
ii = (0,0)
elif i == 1:
ii = (1,0)
elif i ==2:
ii = (1,1)
ax_arr[ii].tick_params(labelsize=14)
ax_arr[ii].plot([0,0],[0,0],'-',color=col_cum[i],markersize=1,label=label_arr[i],linewidth=3.0)
#inc_weights = weights[fits]
import statistics
inc_weights = statistics.renormalise_weights(weights[fits],probs[fits])
inc_weights = inc_weights / np.sum(inc_weights)
prev_weight = np.zeros(len(weights[0]))
tot_weight = np.sum(inc_weights,0)
index_allow = np.where( (age>0) & (metal>-90) )
npts = 300
extent_array = [bounds_interp_age[0],bounds_interp_age[1],bounds_interp_metal[0],bounds_interp_metal[1]]
xage = np.linspace(bounds_interp_age[0],bounds_interp_age[1],npts)
ymetal = np.linspace(bounds_interp_metal[0],bounds_interp_metal[1],npts)
ax_arr[ii].set_xlim(bounds_interp_age[0],bounds_interp_age[1])
ax_arr[ii].set_ylim(bounds_interp_metal[0],bounds_interp_metal[1])
# grid the data.
zi = griddata((age, metal), tot_weight, (xage[None,:], ymetal[:,None]), method='linear')
zi[np.isnan(zi)|np.isinf(zi)]=0.00001
zs =scipy.ndimage.gaussian_filter(zi,sigma=5,order=0)
# im = ax_arr[i].imshow(zi.T[::-1], origin='lower',interpolation='sinc', cmap=cmap,extent=extent_array,\
# vmin=0,vmax=1,aspect='auto')
print np.max(zi)
CS = ax_arr[ii].contour(xage,ymetal,zs/np.max(zs),[0.01,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1.0],\
linewidths=0.5,colors='k',vmin=0.01)
CS = ax_arr[ii].contourf(xage,ymetal,zs/np.max(zs),100,cmap=cmap,vmin=0.01,vmax=1.0)
ax_arr[ii].legend(loc="lower left",prop={'size':18},shadow=False)
# Cumulative plots
age_pdf = np.sum(zi,0)
cum_pdf = np.zeros(np.shape(age_pdf))
for ap in range(len(age_pdf)):
cum_pdf[ap] = np.sum(age_pdf[:ap])
ax_arr[0,1].plot(xage,cum_pdf/np.max(cum_pdf),'-',linewidth=3.0,color=col_cum[i])
ax_arr[0,0].set_title(title,fontsize=20,y=1.05,x=1.0)
ax_arr[0,1].plot([0,0],[0,0],'w-',markersize=0.1,label='CDF',linewidth=3.0)
ax_arr[0,1].legend(loc="upper left",prop={'size':18},shadow=True,handlelength=-0.0,handletextpad=0.0)
ax_arr[0,1].set_xticklabels([""])
ax_arr[0,1].tick_params(labelsize=14)
ax_arr[0,1].set_ylabel("Normalised Weight (< Age)",fontsize=16,rotation=270,labelpad=30)
ax_arr[0,1].yaxis.set_label_position("right")
ax_arr[0,1].yaxis.tick_right()
ax_arr[1,0].set_ylabel("[Z/H]",fontsize=20,y=1.0)
ax_arr[1,0].set_xlabel("Age / log(yr)",fontsize=20,x=1.0)
plt.tight_layout()
f.subplots_adjust(hspace=0,wspace=0)
f.subplots_adjust(bottom=0.2)
cbar_ax = f.add_axes([0.10,0.03,0.8,0.03])#[0.05, 0.15, 0.03, 0.8])
cb = f.colorbar(CS, cax=cbar_ax,cmap=cmap,orientation='horizontal')
cb.cmap.set_under(color='white', alpha=None)
cb.set_ticks([0.01,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1.0])
cb.set_ticklabels(['0.01','0.1','0.2','0.3','0.4','0.5','0.6','0.7','0.8','0.9','1.0'])
cbar_ax.tick_params(length=13, width=1, which='major')
cb.set_label("Relative SSP weight",fontsize=20,labelpad=5)
cb.ax.tick_params(labelsize=18)
out_plot_string = 'plots/contour_'+title+'.eps'
plt.savefig(out_plot_string,format='eps',transparent=True,bbox_inches='tight')
plt.close()
def plot_sfh(age, metal, weights, probs):
# Plot histograms of the SFH of the best fit, 1 sig, (2 sig?) solutions.
# (Maybe with CDF)
# INPUT: ages, metallicities, weights, probs
age = np.log10(age) + 9.0
def diff_array(array):
bisected_array = np.zeros(len(array) - 1)
for ai in range(len(bisected_array)):
bisected_array[ai] = (array[ai + 1] - array[ai])
return bisected_array
min_width = np.min(np.absolute((diff_array(age))))
best_fit = np.argmax(probs)
one_sig_fits = np.where(probs > (1 - 0.6827))
two_sig_fits = np.where(probs > (1 - 0.9545))
fits_arr = [[best_fit], one_sig_fits, two_sig_fits]
print fits_arr
sort_ages = np.sort(age)
sort_age_index = np.argsort(age)
sort_unique_ages = np.sort(np.unique(age))
sort_metals = np.sort(metal)
sort_unique_metals = np.sort(np.unique(metal))
color_metals = [
'violet', 'blue', 'MediumAquaMarine', 'green', 'yellow', 'orange',
'darkred'
]
name_metals = [str(x) for x in sort_unique_metals]
f, ax_arr = plt.subplots(3, sharex=True, sharey=False)
ax_cdf = [ax_arr[0].twinx(), ax_arr[1].twinx(), ax_arr[2].twinx()]
ax_arr[0].plot([-99, -99], [0, 0], 'w.', label='[Z/H] = ')
for nm in range(len(name_metals)):
if name_metals[nm] == '-2.302':
ax_arr[0].bar(-99,
0,
facecolor=color_metals[nm],
label='-2.3 (RHB)')
elif name_metals[nm] == '-2.301':
ax_arr[0].bar(-99,
0,
facecolor=color_metals[nm],
label='-2.3 (BHB)')
elif name_metals[nm] == '-1.302':
ax_arr[0].bar(-99,
0,
facecolor=color_metals[nm],
label='-1.3 (RHB)')
elif name_metals[nm] == '-1.301':
ax_arr[0].bar(-99,
0,
facecolor=color_metals[nm],
label='-1.3 (BHB)')
else:
ax_arr[0].bar(-99,
0,
facecolor=color_metals[nm],
label=name_metals[nm])
for i, fits in enumerate(fits_arr):
#inc_weights = weights[fits]
import statistics
inc_weights = statistics.renormalise_weights(weights[fits],
probs[fits])
inc_weights = inc_weights / np.sum(inc_weights)
prev_weight = np.zeros(len(weights[0]))
tot_weight = np.sum(inc_weights, 0)
print tot_weight
for mi, m in enumerate(sort_unique_metals):
for w in range(len(weights[0])):
if metal[w] == m:
ax_arr[i].bar(age[w],
tot_weight[w],
width=min_width,
facecolor=color_metals[mi],
bottom=prev_weight[w],
align='center')
prev_weight[w] = tot_weight[w] + prev_weight[w]
# Make a CDF:
pdf_bar = np.zeros(len(sort_unique_ages))
cdf_bar = np.zeros(len(sort_unique_ages))
for a in range(len(sort_unique_ages)):
pdf_bar[a] = np.sum(
prev_weight[np.where(age == sort_unique_ages[a])])
cdf_bar[a] = np.sum(pdf_bar[:a])
ax_cdf[i].plot([np.min(sort_unique_ages)-99,np.min(sort_unique_ages)]+sort_unique_ages.tolist()+[np.max(sort_unique_ages),np.max(sort_unique_ages)+99],\
[0,0]+cdf_bar.tolist()+[1.0,1.0],'--',linewidth=4.0,color='springgreen')
# Fine-tune figure; make subplots close to each other and hide x ticks for
# all but bottom plot.c
ax_arr[1].set_ylabel("SSP Weight")
ax_arr[-1].set_xlabel("Age / log(yr)")
ax_cdf[1].set_ylabel("CDF of SSP weight")
handles, labels = ax_arr[0].get_legend_handles_labels()
ax_arr[0].legend(handles,labels,shadow=True,ncol=int(round(len(labels)/2)),bbox_to_anchor=(0.01, 0.95, 0.98, .105), loc=3,\
mode="expand", borderaxespad=0.,handlelength=0.6,handleheight=1.2,handletextpad=0.4,\
framealpha=1.0)
ax_arr[-1].set_xlim(
np.min(age[np.where(prev_weight > 0.001)]) - min_width,
np.max(age[np.where(prev_weight > 0.001)]) + min_width)
f.subplots_adjust(hspace=0)
plt.setp([a.get_xticklabels() for a in ax_arr[:-1]], visible=False)
plt.show()
def plot_sfh_contours(age, metal, weights, probs, title=""):
# Plot FIREFLY'S likelihood-weighted SSP contribution contours
# INPUT: ages, metallicities, weights, probs
age = np.log10(age) + 9.0
metal = np.asarray(metal)
def diff_array(array):
bisected_array = np.zeros(len(array) - 1)
for ai in range(len(bisected_array)):
bisected_array[ai] = (array[ai + 1] - array[ai])
return bisected_array
min_width = np.min(np.absolute((diff_array(age))))
best_fit = np.argmax(probs)
one_sig_fits = np.where(probs > (1 - 0.6827))
two_sig_fits = np.where(probs > (1 - 0.9545))
fits_arr = [[best_fit], one_sig_fits, two_sig_fits]
label_arr = ["Best fit", "68% level", "95% level"]
col_cum = ['black', 'green', 'cyan']
sort_ages = np.sort(age)
sort_age_index = np.argsort(age)
sort_unique_ages = np.sort(np.unique(age))
sort_metals = np.sort(metal)
sort_unique_metals = np.sort(np.unique(metal))
name_metals = [str(x) for x in sort_unique_metals]
f, ax_arr = plt.subplots(2, 2, sharex=True, sharey=True)
# ax_arr = ( (1,2),(3,4) )
def truncate_colormap(cmap, minval=0.0, maxval=1.0, n=100):
new_cmap = colors.LinearSegmentedColormap.from_list(
'trunc({n},{a:.2f},{b:.2f})'.format(n=cmap.name,
a=minval,
b=maxval),
cmap(np.linspace(minval, maxval, n)))
return new_cmap
cmap = truncate_colormap(get_cmap('hot_r'),
minval=0.15,
maxval=1.0,
n=1000)
cmap.set_under('w')
bounds_interp_age = (9.0, 10.19)
bounds_interp_metal = (-2.5, 0.90)
ax_arr[0, 1] = plt.subplot(222)
ax_arr[0, 1].set_xlim(bounds_interp_age[0], bounds_interp_age[1])
#ax_arr[0,1].set_xlabel("Age / log(yr)",fontsize=20)
for i, fits in enumerate(fits_arr):
#ii = (int(i/2),i%2)
if i == 0:
ii = (0, 0)
elif i == 1:
ii = (1, 0)
elif i == 2:
ii = (1, 1)
ax_arr[ii].tick_params(labelsize=14)
ax_arr[ii].plot([0, 0], [0, 0],
'-',
color=col_cum[i],
markersize=1,
label=label_arr[i],
linewidth=3.0)
#inc_weights = weights[fits]
import statistics
inc_weights = statistics.renormalise_weights(weights[fits],
probs[fits])
inc_weights = inc_weights / np.sum(inc_weights)
prev_weight = np.zeros(len(weights[0]))
tot_weight = np.sum(inc_weights, 0)
index_allow = np.where((age > 0) & (metal > -90))
npts = 300
extent_array = [
bounds_interp_age[0], bounds_interp_age[1], bounds_interp_metal[0],
bounds_interp_metal[1]
]
xage = np.linspace(bounds_interp_age[0], bounds_interp_age[1], npts)
ymetal = np.linspace(bounds_interp_metal[0], bounds_interp_metal[1],
npts)
ax_arr[ii].set_xlim(bounds_interp_age[0], bounds_interp_age[1])
ax_arr[ii].set_ylim(bounds_interp_metal[0], bounds_interp_metal[1])
# grid the data.
zi = griddata((age, metal),
tot_weight, (xage[None, :], ymetal[:, None]),
method='linear')
zi[np.isnan(zi) | np.isinf(zi)] = 0.00001
zs = scipy.ndimage.gaussian_filter(zi, sigma=5, order=0)
# im = ax_arr[i].imshow(zi.T[::-1], origin='lower',interpolation='sinc', cmap=cmap,extent=extent_array,\
# vmin=0,vmax=1,aspect='auto')
print np.max(zi)
CS = ax_arr[ii].contour(xage,ymetal,zs/np.max(zs),[0.01,0.1,0.2,0.3,0.4,0.5,0.6,0.7,0.8,0.9,1.0],\
linewidths=0.5,colors='k',vmin=0.01)
CS = ax_arr[ii].contourf(xage,
ymetal,
zs / np.max(zs),
100,
cmap=cmap,
vmin=0.01,
vmax=1.0)
ax_arr[ii].legend(loc="lower left", prop={'size': 18}, shadow=False)
# Cumulative plots
age_pdf = np.sum(zi, 0)
cum_pdf = np.zeros(np.shape(age_pdf))
for ap in range(len(age_pdf)):
cum_pdf[ap] = np.sum(age_pdf[:ap])
ax_arr[0, 1].plot(xage,
cum_pdf / np.max(cum_pdf),
'-',
linewidth=3.0,
color=col_cum[i])
ax_arr[0, 0].set_title(title, fontsize=20, y=1.05, x=1.0)
ax_arr[0, 1].plot([0, 0], [0, 0],
'w-',
markersize=0.1,
label='CDF',
linewidth=3.0)
ax_arr[0, 1].legend(loc="upper left",
prop={'size': 18},
shadow=True,
handlelength=-0.0,
handletextpad=0.0)
ax_arr[0, 1].set_xticklabels([""])
ax_arr[0, 1].tick_params(labelsize=14)
ax_arr[0, 1].set_ylabel("Normalised Weight (< Age)",
fontsize=16,
rotation=270,
labelpad=30)
ax_arr[0, 1].yaxis.set_label_position("right")
ax_arr[0, 1].yaxis.tick_right()
ax_arr[1, 0].set_ylabel("[Z/H]", fontsize=20, y=1.0)
ax_arr[1, 0].set_xlabel("Age / log(yr)", fontsize=20, x=1.0)
plt.tight_layout()
f.subplots_adjust(hspace=0, wspace=0)
f.subplots_adjust(bottom=0.2)
cbar_ax = f.add_axes([0.10, 0.03, 0.8, 0.03]) #[0.05, 0.15, 0.03, 0.8])
cb = f.colorbar(CS, cax=cbar_ax, cmap=cmap, orientation='horizontal')
cb.cmap.set_under(color='white', alpha=None)
cb.set_ticks([0.01, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0])
cb.set_ticklabels([
'0.01', '0.1', '0.2', '0.3', '0.4', '0.5', '0.6', '0.7', '0.8', '0.9',
'1.0'
])
cbar_ax.tick_params(length=13, width=1, which='major')
cb.set_label("Relative SSP weight", fontsize=20, labelpad=5)
cb.ax.tick_params(labelsize=18)
out_plot_string = 'plots/contour_' + title + '.eps'
plt.savefig(out_plot_string,
format='eps',
transparent=True,
bbox_inches='tight')
plt.close()
