def estimateScatteringTimescale(self,
searchtauds=None,
ntauds=25,
name=None,
fourier=False,
**kwargs):
if searchtauds is None:
fwhm = self.getFWHM(timeunits=False) #in bin units
tauds = np.linspace(fwhm / 4, fwhm, ntauds)
tauds = np.linspace(2, self.nbins / 4, ntauds)
tauds = np.linspace(2, fwhm, ntauds)
else:
tauds = np.copy(searchtauds)
bins = np.array(self.bins, dtype=np.float)
N_fs = np.zeros_like(tauds)
sigma_offcs = np.zeros_like(tauds)
Gammas = np.zeros_like(tauds)
f_rs = np.zeros_like(tauds)
for i, taud in enumerate(tauds):
#print i,taud
if fourier:
Dy, C, N_f, sigma_offc, Gamma, f_r = u.pbf_fourier(
bins, self.data, taud=taud, opw=self.opw, **kwargs)
else:
Dy, C, N_f, sigma_offc, Gamma, f_r = u.pbf_clean(bins,
self.data,
taud=taud,
opw=self.opw,
**kwargs)
N_fs[i] = N_f
sigma_offcs[i] = sigma_offc
Gammas[i] = Gamma
f_rs[i] = f_r
f_cs = (Gammas + f_rs) / 2.0
fig = plt.figure(figsize=(8, 12))
ax = fig.add_subplot(511)
ax.plot(tauds, N_fs)
ax.set_ylabel(r'$N_f/N_{\rm tot}$')
ax = fig.add_subplot(512)
ax.plot(tauds, sigma_offcs)
ax.set_ylabel(r'$\sigma_{\rm offc}/\sigma_{\rm off}$')
ax = fig.add_subplot(513)
ax.plot(tauds, Gammas)
ax.set_ylabel(r'$\Gamma$')
ax = fig.add_subplot(514)
ax.plot(tauds, f_rs)
ax.set_ylabel(r'$f_r$')
ax = fig.add_subplot(515)
ax.plot(tauds, f_cs)
ax.set_xlabel(r'$\tau_{\rm d}$')
ax.set_ylabel(r'$f_c$')
if name is None:
plt.show()
else:
np.savez("cleandata/%s_clean_data.npz" % name,
tauds=(tauds * self.getTbin() * 1000),
N_fs=N_fs,
sigma_offcs=sigma_offcs,
Gammas=Gammas,
f_rs=f_rs,
f_cs=f_cs)
plt.savefig("cleanimages/%s_clean_stats.png" % name)
plt.close()
ind = np.argmin(f_cs)
#ind = np.argmin(Gammas)
#print tauds[ind],tauds[ind]*self.getTbin()
if fourier:
Dy, C, N_f, sigma_offc, Gamma, f_r = u.pbf_fourier(bins,
self.data,
taud=tauds[ind],
opw=self.opw,
**kwargs)
else:
Dy, C, N_f, sigma_offc, Gamma, f_r = u.pbf_clean(bins,
self.data,
taud=tauds[ind],
opw=self.opw,
**kwargs)
#Dy,C,N_f,sigma_offc,Gamma,f_r=clean(t,y,taud=10)
#print "gamma",Gamma
fig = plt.figure()
ax = fig.add_subplot(111)
#ax.plot(self.bins,self.data/np.max(self.data),'b')
ax.plot(bins, Dy / np.max(self.data), '0.50')
if fourier:
ax.plot(bins, C / np.max(self.data), 'r')
else:
ax.plot(bins, C, 'r')
ax.plot(bins, self.data / np.max(self.data), 'k')
ax.set_xlim(0, bins[-1])
ax.text(0.7,
0.8,
r'$\tau_d$ = %0.2f ms' % (tauds[ind] * self.getTbin() * 1000),
transform=ax.transAxes,
fontsize=14)
if name is None:
plt.show()
else:
plt.savefig("cleanimages/%s_clean.png" % name)
plt.close()
def estimateScatteringTimescale(self,searchtauds=None,ntauds=25,name=None,fourier=False,**kwargs):
if searchtauds is None:
fwhm = self.getFWHM(timeunits=False) #in bin units
tauds = np.linspace(fwhm/4,fwhm,ntauds)
tauds = np.linspace(2,self.getNbins()/4,ntauds)
tauds = np.linspace(2,fwhm,ntauds)
else:
tauds = np.copy(searchtauds)
bins = np.array(self.bins,dtype=np.float)
N_fs = np.zeros_like(tauds)
sigma_offcs = np.zeros_like(tauds)
Gammas = np.zeros_like(tauds)
f_rs = np.zeros_like(tauds)
for i,taud in enumerate(tauds):
#print i,taud
if fourier:
Dy,C,N_f,sigma_offc,Gamma,f_r = u.pbf_fourier(bins,self.data,taud=taud,opw=self.opw,**kwargs)
else:
Dy,C,N_f,sigma_offc,Gamma,f_r = u.pbf_clean(bins,self.data,taud=taud,opw=self.opw,**kwargs)
N_fs[i] = N_f
sigma_offcs[i] = sigma_offc
Gammas[i] = Gamma
f_rs[i] = f_r
f_cs = (Gammas+f_rs)/2.0
fig = plt.figure(figsize=(8,12))
ax=fig.add_subplot(511)
ax.plot(tauds,N_fs)
ax.set_ylabel(r'$N_f/N_{\rm tot}$')
ax=fig.add_subplot(512)
ax.plot(tauds,sigma_offcs)
ax.set_ylabel(r'$\sigma_{\rm offc}/\sigma_{\rm off}$')
ax=fig.add_subplot(513)
ax.plot(tauds,Gammas)
ax.set_ylabel(r'$\Gamma$')
ax=fig.add_subplot(514)
ax.plot(tauds,f_rs)
ax.set_ylabel(r'$f_r$')
ax=fig.add_subplot(515)
ax.plot(tauds,f_cs)
ax.set_xlabel(r'$\tau_{\rm d}$')
ax.set_ylabel(r'$f_c$')
if name is None:
plt.show()
else:
np.savez("cleandata/%s_clean_data.npz"%name,tauds=(tauds*self.getTbin()*1000),N_fs=N_fs,sigma_offcs=sigma_offcs,Gammas=Gammas,f_rs=f_rs,f_cs=f_cs)
plt.savefig("cleanimages/%s_clean_stats.png"%name)
plt.close()
ind = np.argmin(f_cs)
#ind = np.argmin(Gammas)
#print tauds[ind],tauds[ind]*self.getTbin()
if fourier:
Dy,C,N_f,sigma_offc,Gamma,f_r=u.pbf_fourier(bins,self.data,taud=tauds[ind],opw=self.opw,**kwargs)
else:
Dy,C,N_f,sigma_offc,Gamma,f_r=u.pbf_clean(bins,self.data,taud=tauds[ind],opw=self.opw,**kwargs)
#Dy,C,N_f,sigma_offc,Gamma,f_r=clean(t,y,taud=10)
#print "gamma",Gamma
fig = plt.figure()
ax = fig.add_subplot(111)
#ax.plot(self.bins,self.data/np.max(self.data),'b')
ax.plot(bins,Dy/np.max(self.data),'0.50')
if fourier:
ax.plot(bins,C/np.max(self.data),'r')
else:
ax.plot(bins,C,'r')
ax.plot(bins,self.data/np.max(self.data),'k')
ax.set_xlim(0,bins[-1])
ax.text(0.7,0.8,r'$\tau_d$ = %0.2f ms'%(tauds[ind]*self.getTbin()*1000),transform=ax.transAxes,fontsize=14)
if name is None:
plt.show()
else:
plt.savefig("cleanimages/%s_clean.png"%name)
plt.close()