def p_values_plot(model='Model2',sample='Sample1',test='MannWhitney',show=False,filtered=True,old=False,filename=None,ax=None):
save = ax is None
if ax is None:
fig,ax = plt.subplots(1,1,figsize=size1c)
if filename is None:
filename = 'results/{:s}_{:s}_{:s}.pkl'.format(sample.replace(' ',''),model.replace(' ',''),test)
x,f_false,y,z,dz = load_result(filename)
z,dz = z[:,0,:],dz[:,0,:]
# Gaussian filter for smoother data
if filtered: z,dz = filter_data(z,dz)
ls = ['solid','dashed','dotted']
levels = [0.05,0.1,0.01]
#lws = [lw,1.5*lw,lw/1.5]
lws = [lw,lw,lw]
for i in range(len(levels)):
xs,ys = get_contour(x,y,z,np.log10(levels[i]))
ax.plot(xs,ys,c=cc[1],lw=lws[i],ls=ls[i])
idx = np.argmin(np.abs(xs-0.5))
x0s = [0.475,0.44,0.5]
shifts = [14,-3,25]
rots = [-25,-23,-25]
ax.annotate('p = {:.2f}'.format(levels[i]),xy=(x0s[i],ys[idx]+shifts[i]),color=cc[1],ha='center',va='center',rotation=rots[i])
if i == 0:
# Shaded region marking detectable correlation
ys = ys
ax.fill_between(xs,ys,ys+1e4,color='grey',alpha=0.15,lw=0)
ax.set_xticks([0,0.25,0.5,0.75,1.0])
ax.set_xticklabels(['{:d}%'.format(int(xt*100)) for xt in ax.get_xticks()],y=-0.01)
ax.set_xlim([0,1])
ax.grid(True,axis='y',which='major',lw=2)
ax.grid(True,axis='y',which='minor',lw=0.3)
ax.set_yscale('log')
ax.set_ylim([10,1000])
ax.set_yticks([10,100,1000])
ax.set_yticklabels(['10','100','1000'])
if do_lines:
ax.axvline(0.1,linestyle='dashed',c='black',lw=2)
ax.axvline(0.8,linestyle='dashed',c='black',lw=2)
ax.set_ylabel('Number of exo-Earth candidates',fontsize=lfs)
ax.set_xlabel('Fraction of exo-Earth candidates with life ($f_\mathrm{life}$)',ha='center',va='center',fontsize=lfs,labelpad=30)
plt.subplots_adjust(left=0.13,bottom=0.14,right=0.88)
if show:
plt.show()
elif save:
init_outdir(outdir)
plt.savefig(outdir+'/p_values.pdf')
plt.close()
else:
return ax
def ratio_plots(models=['Model1','Model2','Model3'],samples=['Sample1','Sample2'],test='MannWhitney',ncols=2,show=False,old=False):
# Plots the ratio of the number of planets required for p = 0.05 as a function of f_life between two samples
fig,ax = plt.subplots(1,1,figsize=size1)
# Loop through each model
for i in range(len(models)):
filename1 = 'results/{:s}_{:s}_{:s}.pkl'.format(samples[0].replace(' ',''),models[i].replace(' ',''),test)
filename2 = 'results/{:s}_{:s}_{:s}.pkl'.format(samples[1].replace(' ',''),models[i].replace(' ',''),test)
x1,f_false,y1,z1,dz1 = load_result(filename1,old=old)
x2,f_false,y2,z2,dz2 = load_result(filename2,old=old)
z1,dz1 = z1[:,0,:],dz1[:,0,:]
z2,dz2 = z2[:,0,:],dz2[:,0,:]
if not (np.array_equal(x1,x2) and np.array_equal(y1,y2)):
print("Warning! x/y don't match between samples for {:s}".format(models[i]))
# Get the contour for each
try:
xs1,ys1 = get_contour(x1,y1,z1,np.log10(thresh))
xs2,ys2 = get_contour(x2,y2,z2,np.log10(thresh))
except IndexError:
print("Can't get contour for {:s}".format(models[i]))
continue
# Interpolate onto a common x grid
xs = np.linspace(0,1,100)
ys1 = np.interp(xs,xs1,ys1)
ys2 = np.interp(xs,xs2,ys2)
# Minimum x value
mask = xs>0.15
xs,ys1,ys2 = xs[mask],ys1[mask],ys2[mask]
# Plot the ratio of sample 2 to sample 1
ax.plot(xs,ys2/ys1,c=cc[i],lw=lw,label=models[i]+' ({:.2f})'.format(np.median(ys2/ys1)))
ax.axhline(np.median(ys2/ys1),c=cc[i])
ax.set_title('{:s} / {:s}'.format(sample_names[samples[1]],sample_names[samples[0]]),fontsize=fontsize)
ax.set_xlim([0,1])
ax.set_xticks([0,0.25,0.5,0.75,1.0])
ax.set_xticklabels(['{:d}%'.format(int(xt*100)) for xt in ax.get_xticks()])
ax.legend(loc='best')
plt.show()
def contour_plot(models=['Model1','Model2','Model3'],sample='Sample1',test='MannWhitney',show=False,filtered=True,f_false=0.,old=False,xlog=False,ax=None,fontsize=lfs):
save = ax is None
if ax is None:
fig,ax = plt.subplots(figsize=(1.1*size1c[0],size1c[1]))
ax = [ax]
else:
ax = [ax]
for j in range(len(models)):
filename = 'results/{:s}_{:s}_{:s}.pkl'.format(sample.replace(' ',''),models[j].replace(' ',''),test)
if old:
x,y,z,dz = load_result(filename,old=True)
else:
x,f_falses,y,z,dz = load_result(filename)
# Grab the slice for the correct value of f_false
idx = np.argmin(np.abs(f_false-f_falses))
z,dz = z[:,idx,:],dz[:,idx,:]
if np.abs(f_false-f_falses[idx]) > 0.01:
print("Warning: closest match to f_false = {:.4f} in the results pkl is f_false = {:.4f}".format(f_false,f_falses[idx]))
# Gaussian filter for smoother data
if filtered: z,dz = filter_data(z,dz)
xs,ys = get_contour(x,y,z,np.log10(thresh))
col = cc[j]
ax[0].plot(xs,ys,c=col,lw=lw)
x0 = [0.45,0.67,0.53]
y0 = [0.2,0.152,0.34]
rot = [-27,-23,-23]
ax[0].annotate(model_names[models[j]],xy=(x0[j],y0[j]),xycoords='axes fraction',color=col,ha='center',va='center',
rotation=rot[j],zorder=100,weight='bold' if j == 1 else 'normal',fontsize=0.9*plt.rcParams['font.size'])
if j == 1:
# Shaded region marking detectable correlation
ys = ys
ax[0].fill_between(xs,ys,ys+1e4,color='grey',alpha=0.15,lw=0)
ax[0].annotate('Detectable correlation',xy=(0.5,0.69),xycoords='axes fraction',ha='center',va='center',color='grey',rotation=0)
if xlog:
ax[0].set_xscale('log')
else:
ax[0].set_xticks([0,0.25,0.5,0.75,1.0])
ax[0].set_xticklabels([' {:d}%'.format(int(xt*100)) for xt in ax[0].get_xticks()],y=-0.01)
ax[0].set_xlim([0,1])
ax[0].grid(True,axis='y',which='major',lw=2)
ax[0].grid(True,axis='y',which='minor',lw=0.3)
ax[0].set_yscale('log')
ax[0].set_ylim([10,1000])
ax[0].set_yticks([10,100,1000])
ax[0].set_yticklabels(['10','100','1000'])
ax[0].axvline(0.1,linestyle='dashed',c='black',lw=2)
ax[0].axvline(0.8,linestyle='dashed',c='black',lw=2)
ax[0].annotate('Optimistic',xy=(0.81,950),rotation=90,va='top',ha='left',fontsize=1.25*afs)
ax[0].annotate('Pessimistic',xy=(0.11,950),rotation=90,va='top',ha='left',fontsize=1.25*afs)
# Estimates for LUVOIR, Nautilus, LIFE, Origins
obs_names = ['LUVOIR','Nautilus Space\nObservatory','LIFE','Origins Space\nTelescope']
obs_vals = [55.5,986,44.5,26]
ax2 = ax[0].twinx()
ax2.set_ylim([1,3])
ax2.set_yticks(np.log10(obs_vals))
ax2.set_yticklabels(obs_names,fontsize=1.25*afs)
ax2.tick_params(axis='y',width=4,length=15,pad=15)
ax[0].set_ylabel('Number of\nexo-Earth candidates',fontsize=fontsize)
ax[0].set_xlabel('Fraction of exo-Earth candidates with life ($f_\mathrm{life}$)',ha='center',va='center',fontsize=fontsize,labelpad=30)
plt.subplots_adjust(left=0.22,bottom=0.14,right=0.78)
if show:
plt.show()
elif save:
init_outdir(outdir)
plt.savefig(outdir+'/Figure_2.pdf')
plt.close()
else:
return ax[0]
def sample_selection_plot(model='Model2',samples=['Sample1','Sample2','Sample4'],test='MannWhitney',show=False,filtered=True,old=False,ax=None):
save = ax is None
if ax is None:
fig,ax = plt.subplots(1,1,figsize=size1c)
for i in range(len(samples)):
filename = 'results/{:s}_{:s}_{:s}.pkl'.format(samples[i].replace(' ',''),model.replace(' ',''),test)
x,f_false,y,z,dz = load_result(filename)
z,dz = z[:,0,:],dz[:,0,:]
# Gaussian filter for smoother data
if filtered: z,dz = filter_data(z,dz)
ls = ['solid','dashed','dotted']
labels = ['All planets','Young and old planets only','No young planets']
xs,ys = get_contour(x,y,z,np.log10(thresh))
ax.plot(xs,ys,c=cc[1],lw=lw,ls=ls[i])
idx = np.argmin(np.abs(xs-0.45))
shifts = [11,-1.75,5]
xshifts = [0,-.085,.08]
rots = [-25,-24,-25]
ax.annotate(labels[i],xy=(0.45+xshifts[i],ys[idx]+shifts[i]),color=cc[1],ha='center',va='center',rotation=rots[i])
if i == 0:
# Shaded region marking detectable correlation
ys = ys
ax.fill_between(xs,ys,ys+1e4,color='grey',alpha=0.15,lw=0)
ax.annotate('Detectable correlation',xy=(0.95,0.92),xycoords='axes fraction',ha='right',va='center',color='grey',rotation=0,weight='bold',
bbox=dict(boxstyle='round',fc='grey',lw=0,alpha=0.05))
ax.set_xticks([0,0.25,0.5,0.75,1.0])
ax.set_xticklabels(['{:d}%'.format(int(xt*100)) for xt in ax.get_xticks()],y=-0.01)
ax.set_xlim([0,1])
ax.grid(True,axis='y',which='major',lw=2)
ax.grid(True,axis='y',which='minor',lw=0.3)
ax.set_yscale('log')
ax.set_ylim([10,1000])
ax.set_yticks([10,100,1000])
ax.set_yticklabels(['10','100','1000'])
if do_lines:
ax.axvline(0.1,linestyle='dashed',c='black',lw=2)
ax.axvline(0.8,linestyle='dashed',c='black',lw=2)
ax.set_ylabel('Number of exo-Earth candidates',fontsize=lfs)
ax.set_xlabel('Fraction of exo-Earth candidates with life ($f_\mathrm{life}$)',ha='center',va='center',fontsize=lfs,labelpad=30)
plt.subplots_adjust(left=0.13,bottom=0.14,right=0.88)
if show:
plt.show()
elif save:
init_outdir(outdir)
plt.savefig(outdir+'/sample_selection.pdf')
plt.close()
else:
return ax
def test_comparison_plot(model='Model2',sample='Sample1',tests=['MannWhitney','Spearman','Student'],show=False,filtered=True,ax=None):
save = ax is None
if ax is None:
fig,ax = plt.subplots(1,1,figsize=size1c)
for i in range(len(tests)):
filename = 'results/{:s}_{:s}_{:s}.pkl'.format(sample.replace(' ',''),model.replace(' ',''),tests[i])
x,f_false,y,z,dz = load_result(filename)
z,dz = z[:,0,:],dz[:,0,:]
# Gaussian filter for smoother data
if filtered: z,dz = filter_data(z,dz)
ls = ['solid','dashed','dotted','dashdot']
label = test_names[tests[i]]
xs,ys = get_contour(x,y,z,np.log10(thresh))
ax.plot(xs+(0.005 if i == 2 else 0),ys,c=cc[1],lw=lw,ls=ls[i])
idx = np.argmin(np.abs(xs-0.5))
if i == 0:
# Shaded region marking detectable correlation
ys = ys
ax.fill_between(xs,ys,ys+1e4,color='grey',alpha=0.15,lw=0)
for i in range(2):
x0 = [0.75,0.45]
y0 = [0.21,0.243]
rots = [-18.,-24]
labels = ['Mann-Whitney U test',"Spearman's / Student's tests"]
ax.annotate(labels[i],xy=(x0[i],y0[i]),color=cc[1],xycoords='axes fraction',ha='center',va='center',rotation=rots[i])
ax.set_xticks([0,0.25,0.5,0.75,1.0])
ax.set_xticklabels(['{:d}%'.format(int(xt*100)) for xt in ax.get_xticks()],y=-0.01)
ax.set_xlim([0,1])
ax.grid(True,axis='y',which='major',lw=2)
ax.grid(True,axis='y',which='minor',lw=0.3)
ax.set_yscale('log')
ax.set_ylim([10,1000])
ax.set_yticks([10,100,1000])
ax.set_yticklabels(['10','100','1000'])
if do_lines:
ax.axvline(0.1,linestyle='dashed',c='black',lw=2)
ax.axvline(0.8,linestyle='dashed',c='black',lw=2)
ax.set_ylabel('Number of exo-Earth candidates',fontsize=lfs)
ax.set_xlabel('Fraction of exo-Earth candidates with life ($f_\mathrm{life}$)',ha='center',va='center',fontsize=lfs,labelpad=30)
plt.subplots_adjust(left=0.13,bottom=0.14,right=0.88)
if show:
plt.show()
elif save:
init_outdir(outdir)
plt.savefig(outdir+'/test_comparison.pdf')
plt.close()
else:
return ax