def plot_fits(mira):
## get the data
star = np.loadtxt(mira,
usecols=(0,1,2),skiprows=0)
ctimes = star[star[:,1] > 29.5,0]
times = star[star[:,1] < 29.5,0]
mags = star[star[:,1] < 29.5,1]
errors = star[star[:,1] < 29.5,2]
## get lomb-scargle period
freqs = lomb.get_freqs2(times)
rss = lomb.lomb(times,mags,errors,freqs)
period = 1. / freqs[np.argmin(rss)]
## from lomb, get A,B,C values
sin_c, cos_c = likelihood.times2pred(ctimes,1./period)
sin_uc, cos_uc = likelihood.times2pred(times,1./period)
f = likelihood.mag2flux(mags)
e = likelihood.magerr2fluxerr(mags,errors)
D = np.array([ np.sin(2*np.pi*times/period),
np.cos(2*np.pi*times/period),
np.ones(times.size)])
A,B,C = np.linalg.lstsq(D.T,mags)[0]
## fit censoring model to data
pars = np.array((A,B,C))
mu_b = np.min(f)
v_b = 10.0
cens = False
pars_fit = op.fmin(likelihood.nll_fixedB_no_cens,pars,
(mu_b,v_b,sin_c,cos_c,sin_uc,cos_uc,f,e))
## set x limits to time min and max
## add legend
## data for plotting
times_pred = np.linspace(np.min(times),np.max(times),1000)
mags_pred_ls = (A*np.sin(2.*np.pi*times_pred / period)
+ B*np.cos(2.*np.pi*times_pred / period)
+ C)
mags_pred_cens = (pars_fit[0]*np.sin(2.*np.pi*times_pred / period)
+ pars_fit[1]*np.cos(2.*np.pi*times_pred / period)
+ pars_fit[2])
xmin = np.min(times)
xmax = np.max(times)
fig = plt.figure()
leg_font = matplotlib.font_manager.FontProperties()
leg_font.set_size(10)
## plot in mag space
ax = fig.add_subplot(211)
line1 = ax.plot(times_pred,-mags_pred_ls,color='orange',linewidth=1.5)
line2 = ax.plot(times_pred,-mags_pred_cens,color='blue',linewidth=1.5)
ax.plot(ctimes,np.min(-mags)*np.ones(ctimes.size),'o',color='red')
ax.plot(times,-mags,'o',color="gray",alpha=.5)
ax.set_yticklabels(np.abs(ax.get_yticks()))
ax.axis([xmin,xmax,ax.axis()[2],ax.axis()[3]])
plt.legend((line1,line2),("Lomb-Scargle","Censored"),loc='upper left',
prop=leg_font)
ax.set_ylabel("Magnitude")
ax.set_xlabel("Time")
## plot in flux space
ax = fig.add_subplot(212)
line1 = ax.plot(times_pred,likelihood.mag2flux(mags_pred_ls),
color='orange',linewidth=1.5)
line2 = ax.plot(times_pred,likelihood.mag2flux(mags_pred_cens),
color='blue',linewidth=1.5)
ax.plot(ctimes,likelihood.mag2flux(np.max(mags))*np.ones(ctimes.size),
'o',color='red')
ax.plot(times,likelihood.mag2flux(mags),'o',color="gray",alpha=.5)
ax.axis([xmin,xmax,ax.axis()[2],ax.axis()[3]])
ax.set_ylabel("Flux")
ax.set_xlabel("Time")
plt.legend((line1,line2),("Lomb-Scargle","Censored"),loc='upper left',
prop=leg_font)
mira_name = mira.split('/')[-1][:-4]
plt.savefig("diag_figs/mira_ls_censor_simple/" + mira_name + ".pdf")
plt.close()
ax = fig.add_subplot(111)
ax.plot(t,likelihood.pred2mu(sins,coss,A,B,C))
plt.savefig("diag_figs/times2pred_and_pred2mu.pdf")
plt.close()
## CHECK: mag2flux and flux2 mag
t = np.linspace(0,10,1000)
sins, coss = likelihood.times2pred(t,w=.7)
A = 1.5
B = 1.3
C = 7.
mu = likelihood.pred2mu(sins,coss,A,B,C)
f = likelihood.mag2flux(mu)
mu2 = likelihood.flux2mag(f)
# first and third plots should be the same
fig = plt.figure()
ax = fig.add_subplot(311)
ax.plot(t,mu)
ax2 = fig.add_subplot(312)
ax2.plot(t,f)
mu = likelihood.flux2mag(f)
ax3 = fig.add_subplot(313)
ax3.plot(t,mu2)
plt.savefig("diag_figs/mag2flux_and_flux2.pdf")
plt.close()