def plotzs(template_file, msfiles, flips):
'''Designed to plot line profiles at the same radius from various heights.
template_file should be .ms-like (either .ms or bin.ms) and will be used
to construct the bins that will be extracted from all other msfiles
'''
template_slay = template_file.split('.ms')[0]+'.slay.fits'
print template_slay
kpcradii, _, _ = sa.openslay(template_slay,flip=flips[0])
HDU = pyfits.open(template_file)[0]
head = HDU.header
bins = []
i = 1
while 'APNUM{}'.format(i) in head:
rstr = head['APNUM{}'.format(i)].split(' ')
bins.append([int(rstr[2]),int(rstr[3])])
i += 1
print len(bins), kpcradii.size
widths = np.diff(np.array(bins)).flatten() * 0.118*8*34.1e3/206265
for (msfile,flip) in zip(msfiles + [template_file],flips):
name = msfile.split('.ms.fits')[0] + '_binplot.pdf'
pp = PDF(name)
tempr, _, _ = sa.openslay(msfile.split('.ms')[0] + '.slay.fits',
flip=flip)
tmphead = pyfits.open(msfile)[0].header
tmprstr = tmphead['APNUM1'].split(' ')
tmpwidth = int(tmprstr[3]) - int(tmprstr[2])
tmpwidth *= 0.118*8*34.1e3/206265
for radius in kpcradii:
idx = np.where(np.abs(tempr-radius) == np.min(np.abs(tempr-radius)))[0][0]
tmprstr = tmphead['APNUM{}'.format(idx+1)].split(' ')
print tmprstrx
dpx = int(tmprstr[3]) - int(tmprstr[2])
dr = dpx * 0.118*8*34.1e3/206265
r = tempr[idx]
fig0= plt.figure()
ax0 = fig0.add_subplot(111)
ax0.set_title('{:}\nr = {:5.3f} kpc\ndr = {:5.3f} kpc'.\
format(msfile,r,dr))
ax0.set_xlabel('Wavelength [A]')
ax0.set_ylabel('ADU/s')
sa.plot_line(msfile,r,ax=ax0,plot=False,flip=flip)
pp.savefig(fig0)
pp.close()
return
def test_kurt(slayfile,outname):
radii, _, _ = sa.openslay(slayfile,moments=False)
zlist = [0,1,2,4]
for z in zlist:
simfile = 'sim_z{}_boring.fits'.format(z)
flarefile = 'sim_z{}_nflarea_1245.fits'.format(z)
ax = plt.figure().add_subplot(111)
ax.set_title('Generated on {}'.format(datetime.now().isoformat()))
ax.set_xlabel('Radius [km/s]')
ax.set_ylabel('$\mu_4$')
ax.set_title('z = {}$h_z$'.format(z))
bkurt = np.array([])
fkurt = np.array([])
for radius in radii:
_, bmV, bmI, _ = do_line(simfile, radius, 1.,
plot=False,rwidth=9.3)
_, fmV, fmI, _ = do_line(flarefile, radius, 1.,
plot=False,rwidth=9.3)
bmoments = ADE.ADE_moments(bmV, bmI)
fmoments = ADE.ADE_moments(fmV, fmI)
bkurt = np.append(bkurt, bmoments[3])
fkurt = np.append(fkurt, fmoments[3])
ax.plot(radii, bkurt, label='simple disk')
ax.plot(radii, fkurt, label='flare')
ax.legend(loc=0)
pp = PDF('{}_z{}.pdf'.format(outname,z))
pp.savefig(ax.get_figure())
pp.close()
return
def MGD(slayfile,radius,cent_lambda=5048.126,
flip=False,window=15,tol=3.,baseline=1):
V, I, err, rwidth = sa.plot_line(slayfile,radius,wavelength=cent_lambda,
velo=True,plot=False,baseline=baseline,
window=window,flip=flip)
factor = 1e6
VV = V*factor
II = np.copy(I)
II[II < 0] = 0.0
II /= np.sum(II)
line_pdf = ss.rv_discrete(a=VV.min(),b=VV.max(),values=(VV,II))
line_sample = line_pdf.rvs(size=10000)/factor
N = np.arange(1,5)
models = [GMM(n).fit(line_sample) for n in N]
AIC = [m.aic(line_sample) for m in models]
BIC = [m.bic(line_sample) for m in models]
M_best = models[np.argmin(AIC)]
clean_model(M_best,tol=tol)
mV = np.linspace(V.min(),V.max(),1000.)
lp, resp = M_best.score_samples(mV)
model_pdf = np.exp(lp)
model_components = resp * model_pdf[:,np.newaxis]
fig = plt.figure()
ax1 = fig.add_subplot(311)
ax1.plot(mV,model_pdf,'k')
ax1.plot(mV,model_components,'--k')
ax1.errorbar(V,I/np.max(I)*np.max(model_pdf),
yerr=err/np.max(I)*np.max(model_pdf),alpha=0.4)
ax1.hist(line_sample,30,normed=True,alpha=0.8,histtype='step')
ax1.set_xlabel('Velocity [km/s]')
ax1.set_ylabel('Normalized counts')
ax2 = fig.add_subplot(323)
ax2.plot(N,BIC,'--k',label='BIC')
ax2.plot(N,AIC,'-k',label='AIC')
ax2.legend(loc=0)
ax2.set_xlabel('Number of Gaussian components')
ax2.set_ylabel('Information Criterion')
# fig.suptitle('r = {:5.3f} kpc\n{:}'.format(radius,time.asctime()))
# fig.show()
return mV, model_pdf, V, I, err, rwidth, fig
def grid_plot(msfiles, flips, sims, simleg):
font = {'size':4}
rc('font',**font)
for msfile, flip, simfiles in zip(msfiles,flips,sims):
print msfile
height = int(msfile.split('_')[1][1:])
if height == 5:
height = 0.5
name = msfile.split('.ms.fits')[0] + '_binplot.pdf'
simname = '_'.join(simfiles[0].split('_')[0:2]) + '_binplot.pdf'
pp = PDF(name)
simpp = PDF(simname)
tempr, _, _ = sa.openslay(msfile.split('.ms')[0] + '.slay.fits',
flip=flip)
print tempr
tmphead = pyfits.open(msfile)[0].header
ap = 1
fig0 = plt.figure(figsize=(11,8.5))
fig1 = plt.figure(figsize=(11,8.5))
for radius in np.sort(tempr):
print '\t{}'.format(radius)
tmprstr = tmphead['APNUM{}'.format(ap)].split(' ')
tmpwidth = int(tmprstr[3]) - int(tmprstr[2])
tmpwidth *= 0.118*8*34.1e3/206265
#############
ax0 = fig0.add_subplot(3,4,ap)
ax0.set_title('{}\n{}'.\
format(msfile,datetime.now().isoformat(' ')),
fontsize=4)
ax0.set_xlabel('Velocity [km/s]')
ax0.set_ylabel('ADU/s')
ax0.text(0.1,0.95,'z ~ {:}$h_z$\nr = {:5.3f}kpc\ndr = {:5.3f}kpc'\
.format(height,radius,tmpwidth),
transform=ax0.transAxes,
ha='left',va='top')
ax0.set_xlim(-600,600)
sa.plot_line(msfile,radius,ax=ax0,plot=False,flip=flip,velo=True,
baseline=1,linewidth=0.4)
#############
ax1 = fig1.add_subplot(3,4,ap)
ax1.set_title('{}\n{}'.\
format(simname.split('_binplot')[0]
,datetime.now().isoformat(' ')),
fontsize=4)
ax1.text(0.1,0.65,'z = {:}$h_z$\nr = {:5.3f}kpc\ndr = {:5.3f}kpc'\
.format(height,radius,tmpwidth),
transform=ax1.transAxes,
ha='left',va='top')
ax1.set_xlabel('Velocity [km/s]')
ax1.set_ylabel('Normalized flux')
ax1.set_xlim(-600,600)
'''The negative 1 is there because all my sims are created
backwards, by convention
'''
for simfile in simfiles:
variable = pyfits.open(simfile)[0].header[simleg.upper()]
mt.do_line(simfile,-1*radius,1,ax=ax1,plot=False,
label=str(variable),rwidth=tmpwidth,linewidth=0.8)
ax1.legend(loc=0,title=simleg)
ap += 1
pp.savefig(fig0)
pp.close()
simpp.savefig(fig1)
simpp.close()
return
def get_drunk(slayfile, baseoutput, flip=False, cdf_window=0.05, tol=3.,
window=15,baseline=1,skip_radii=[],cent_lambda=5048.126,interact=False):
fitsout = baseoutput+'.fits'
pdfout = baseoutput+'.pdf'
pp = PDF(pdfout)
radii, _, _ = sa.openslay(slayfile,flip=flip)
outradii = np.array([])
widths = np.array([])
vwidths = np.array([])
m1 = np.empty((2,1))
m2 = np.empty((2,1))
m3 = np.empty((2,1))
for i, radius in enumerate(radii):
if int(np.floor(radius)) in skip_radii:
print "user skipping radius {} kpc".format(radius)
continue
print 'Getting r = {:5.3} kpc'.format(radius)
moments, moment_err, rwidth, vwidth, fig = cocaine(slayfile, radius,
flip=flip,tol=tol,
cdf_window=cdf_window,
baseline=baseline,
window=window,
cent_lambda=cent_lambda,
interact=interact)
if np.isnan(np.sum(moments)):
print 'skipping radius {} kpc due to NaN condition'.format(radius)
continue
m1 = np.hstack((m1,np.array([[moments[0],moment_err[0]]]).T))
m2 = np.hstack((m2,
np.array([[np.sqrt(moments[1]),
moment_err[1]/(2*np.sqrt(moments[1]))]]).T))
m3 = np.hstack((m3,np.array([[moments[2],moment_err[2]]]).T))
widths = np.append(widths,rwidth)
vwidths = np.append(vwidths,vwidth)
outradii = np.append(outradii,radius)
fig.tight_layout(pad=0.5)
fig.suptitle('r = {:5.3f} kpc\n{:}'.format(radius,time.asctime()))
pp.savefig(fig)
m1 = m1[:,1:]
m2 = m2[:,1:]
m3 = m3[:,1:]
radiiHDU = pyfits.PrimaryHDU(np.vstack((outradii,widths,vwidths)))
radiiHDU.header.update('SLAY',slayfile,comment='Slay file')
radiiHDU.header.update('DATE',time.asctime(),comment='Date of extraction')
radiiHDU.header.update('FLIP',flip,comment='Flip radii around 0?')
radiiHDU.header.update('CDF_WIN',cdf_window,comment='Limits on CDF window')
radiiHDU.header.update('TOL',tol,comment='Gaussian rejection tolerance')
radiiHDU.header.update('EXTNAME','Radii')
m1HDU = pyfits.ImageHDU(m1)
m1HDU.header.update('EXTNAME','mu1')
m2HDU = pyfits.ImageHDU(m2)
m2HDU.header.update('EXTNAME','sqrt(mu2)')
m3HDU = pyfits.ImageHDU(m3)
m3HDU.header.update('EXTNAME','mu3')
pyfits.HDUList([radiiHDU,m1HDU,m2HDU,m3HDU]).writeto(fitsout,clobber=True)
pp.close()
plt.close('all')
return
def moments_notice(slayfile, simfile, plotprefix=False,nofits=False,
flip=False, cent_lambda = 5048.126, skip_radii = []):
'''Take an extracted spectrum (slayfile) and model galaxy (simfile) and
compute the first 3 statistical moments for both at all radii sampled by
the spectrum. The model is binned and degraded to match the data quality
of the actual data. The return values are the radii used, and then a tuple
for each of the first three moments containing the moment and error on the
data and the moment of the model. The moments are computed in the same
window for both model and data so that we can compare the two as closely
as possible. The exception is the first moment.
'''
# radii, rwidths, vwidths, m1, m2, m3 = drunkdata
radii, _, _ = sa.openslay(slayfile,flip=flip,moments=False)
big_dm1 = np.array([])
big_dm1e = np.array([])
big_dm2 = np.array([])
big_dm2e = np.array([])
big_dm3 = np.array([])
big_dm3e = np.array([])
big_mm1 = np.array([])
big_mm2 = np.array([])
big_mm3 = np.array([])
plot_radii = np.array([])
if plotprefix:
pp = PDF(plotprefix+'_lines.pdf')
for radius in radii:
if int(np.floor(radius)) in skip_radii:
print "user skipping radius {} kpc".format(radius)
continue
dV, dI, derr, rwidth = sa.plot_line(slayfile,radius,verbose=False,
wavelength=cent_lambda,velo=True,
plot=False,baseline=1,flip=flip)
mV, mI, _ = salty.line_profile(simfile,radius,plot=False,Iwidth=17,
width=rwidth,observe=True,fit=False,
verbose=False,nofits=nofits)
'''We need to compute the peak of the model line separately because we
shifted it to find the window used for the higher order moments'''
mpeak, mwidth, _, _ = ADE.ADE_moments(mV,mI)
vwidth = mwidth**0.5
mlowV = mpeak - vwidth/2.
mhighV = mpeak + vwidth/2.
mmoment_idx = np.where((mV >= mlowV) & (mV <= mhighV))
mmoments = ADE.ADE_moments(mV[mmoment_idx],mI[mmoment_idx])
'''Now find the peak in the data'''
dpeak, _, _, _ = ADE.ADE_moments(dV,dI)
dlowV = dpeak - vwidth/2.
dhighV = dpeak + vwidth/2.
dmoment_idx = np.where((dV >= dlowV) & (dV <= dhighV))
dmoments, derr = ADE.ADE_moments(dV[dmoment_idx],dI[dmoment_idx],err=derr[dmoment_idx])
if dmoment_idx[0].size == 0 or mmoment_idx[0].size ==0:
continue
# print mmoments
big_mm1 = np.append(big_mm1,mmoments[0])
big_mm2 = np.append(big_mm2,np.sqrt(mmoments[1]))
big_mm3 = np.append(big_mm3,mmoments[2])
big_dm1 = np.append(big_dm1,dmoments[0])
big_dm1e = np.append(big_dm1e,derr[0])
big_dm2 = np.append(big_dm2,dmoments[1])
big_dm2e = np.append(big_dm2e,derr[1])
big_dm3 = np.append(big_dm3,dmoments[2])
big_dm3e = np.append(big_dm3e,derr[2])
plot_radii = np.append(plot_radii,radius)
return plot_radii, [big_dm1, big_dm1e, big_mm1],\
[big_dm2, big_dm2e, big_mm2],\
[big_dm3, big_dm3e, big_mm3]
def line_comp(simfile,slayfile,plotprefix,width=17,flip=False):
'''compares simulation and actual data on line shapes, using a moment based
approach'''
pp = PDF(plotprefix+'_lines.pdf')
# ppg = PDF(plotprefix+'_linegrid.pdf')
gridfig = plt.figure()
dradii, dcent, derr, dm1, dm2, dm3 = sa.openslay(slayfile,moments=True,
flip=flip)
print dradii
rwidth = np.abs(np.mean(np.diff(dradii)))
tm1 = np.array([])
tm2 = np.array([])
tm3 = np.array([])
central_lambda=np.array([4901.416,5048.126])
print "{:^20} {:^20} {:^20}\n{:^10}{:^10}".format('m1','m2','m3','model','data')
print ("-"*20+' ')*3
for i in range(dradii.size):
radius = dradii[i]
(mm1, mm2, mm3), V, I, _ = do_line(simfile,radius,1.,
plot=False,Iwidth=width,rwidth=rwidth)
tm1 = np.append(tm1,mm1)
tm2 = np.append(tm2,mm2)
tm3 = np.append(tm3,mm3)
# (nm1,nm2,nm3), negV, negI, _ = do_line(simfile,-1*radius,1.,
# plot=False,Iwidth=width)
# cent_lambda = central_lambda[np.where(
# np.abs(central_lambda - dm1[i,1]) == np.min(
# np.abs(central_lambda - dm1[i,1])))]
if np.isnan(dcent[i]):
continue
cent_lambda = 5048.126
plot_center = dcent[i]/(3.e5) * cent_lambda + cent_lambda
wave, spec, err = sa.plot_line(slayfile,radius,
wavelength=plot_center,plot=False)
spec -= np.mean(spec[0:10])
dvelo = (wave - cent_lambda)/cent_lambda*3.e5
if flip:
dvelo *= -1.0
fig0 = plt.figure()
ax0 = fig0.add_subplot(111)
ax0.errorbar(dvelo,spec/np.max(spec),yerr=err/np.max(spec),color='gray')
ax0.plot(V,I/np.max(I),color='b')
#ax0.plot(negV+mm1-nm1,negI/np.max(negI))
ax0.set_xlabel('Velocity [km/s]')
ax0.set_ylabel('Signal')
ax0.set_title(datetime.now().isoformat(' ')+'\nr = {:4.3f} kpc'.
format(dradii[i]))
print "{:10.3f} {:10.3f} {:10.3f} {:10.3f} {:10.3f} {:10.3f}".format(
mm1,dm1[i,1],mm2,dm2[i,1],mm3,dm3[i,1])
pp.savefig(fig0)
# ax0.change_geometry(4,5,i+1)
# gridfig.axes.append(ax0)
# fig0.clf()
pp.close()
pp1 = PDF(plotprefix+'_moments.pdf')
fig1 = plt.figure()
ax1 = fig1.add_subplot(111)
ax1.plot(dradii,tm1,'.',label='model')
ax1.plot(dradii,dcent,'x',label='data centers')
ax1.plot(dradii,dm1[:,1],'x',label='data $\mu_1$')
ax1.set_xlabel('Radius [kpc]')
ax1.set_ylabel('$\mu_1$')
ax1.set_ylim(-260,260)
ax1.legend(loc=0,scatterpoints=1,numpoints=1)
ax1.set_title(datetime.now().isoformat(' '))
pp1.savefig(fig1)
fig2 = plt.figure()
ax2 = fig2.add_subplot(111)
ax2.plot(dradii,tm2,'.',label='model')
ax2.plot(dradii,dm2[:,1],'x',label='data')
ax2.set_xlabel('Radius [kpc]')
ax2.set_ylabel('$\mu_2$')
ax2.set_ylim(0,2000)
ax2.legend(loc=0,scatterpoints=1,numpoints=1)
pp1.savefig(fig2)
fig3 = plt.figure()
ax3 = fig3.add_subplot(111)
ax3.plot(dradii,tm3,'.',label='model')
# ax3.plot(dradii,dm3[:,1],'x',label='data')
ax3.set_xlabel('Radius [kpc]')
ax3.set_ylabel('$\mu_3$')
ax3.set_ylim(-2,2)
ax3.legend(loc=0,scatterpoints=1,numpoints=1)
pp1.savefig(fig3)
pp1.close()
# ppg.savefig(gridfig)
# ppg.close()
plt.clf()
# plt.close('all')
return
def new_moments(slayfile, simfile, plotprefix, flip=False, cent_lambda = 5048.126):
radii, _, _ = sa.openslay(slayfile,flip=flip,moments=False)
# rwidth = np.mean(np.diff(radii))
big_dm3 = np.array([])
big_mm3 = np.array([])
big_em3 = np.array([])
big_mm3_2 = np.array([])
plot_radii = np.array([])
if plotprefix:
pp = PDF(plotprefix+'_lines.pdf')
for radius in radii:
dV, dI, derr, rwidth = sa.plot_line(slayfile,radius,wavelength=cent_lambda,velo=True,
plot=False,baseline=1,flip=flip)
(_, _, mm3,_), mV, mI, _ = do_line(simfile, radius, 1.,
plot=False,rwidth=rwidth)
conv_dI = dI/np.mean(dI)
conv_mI = mI/np.mean(mI)
mI_pad = pad(mI,dI.size)
corr = np.convolve(conv_dI,mI_pad[::-1],'same')
idx = np.where(corr == corr.max())[0][0]
peak = dV[idx]
fit_mV = dV + peak
cdf = np.cumsum(mI_pad/np.sum(mI_pad))
# Limits defined by model data
lowV = np.interp(0.1,cdf,fit_mV)
highV = np.interp(0.9,cdf,fit_mV)
dmoment_idx = np.where((dV >= lowV) & (dV <= highV))
mmoment_idx = np.where((fit_mV >= lowV) & (fit_mV <= highV))
print dmoment_idx
# Limits defined by smoothed actual data
window_size = 21
window = np.blackman(window_size)
signal = np.r_[dI[window_size - 1:0:-1],dI,dI[-1:-1*window_size:-1]]
smoothed = np.convolve(window/np.sum(window),signal,'valid')[(window_size/2):-(window_size/2)]
scdf = np.cumsum(smoothed/np.sum(smoothed))
dlowV = np.interp(0.01,scdf,dV)
dhighV = np.interp(0.99,scdf,dV)
dsmooth_idx = np.where((dV >= dlowV) & (dV <= dhighV))
if plotprefix:
fig = plt.figure()
ax0 = fig.add_subplot(311)
ax1 = fig.add_subplot(312)
ax2 = fig.add_subplot(313)
ax0.set_title('r = {:5.3f} kpc'.format(radius))
ax0.plot(dV,dI,'k-')
ax0.plot(mV,mI*np.max(dI)/np.max(mI),'b:',alpha=0.6)
ax0.plot(fit_mV,mI_pad*np.max(dI)/np.max(mI_pad),'b-')
ax0.axvline(x=lowV,ls=':',color='r',alpha=0.8)
ax0.axvline(x=highV,ls=':',color='r',alpha=0.8)
ax0.axvline(x=dlowV,ls=':',color='g',alpha=0.8)
ax0.axvline(x=dhighV,ls=':',color='g',alpha=0.8)
ax1.plot(dV,corr)
ax1.axvline(x=peak,ls=':',color='k',alpha=0.5)
ax2.plot(dV,scdf)
ax2.set_xlim(ax0.get_xlim())
ax1.set_xlim(ax0.get_xlim())
pp.savefig(fig)
plt.close(fig)
if dmoment_idx[0].size <= 2:
print "skipping {} kpc".format(radius)
continue
# _, _, dm3 = ADE.ADE_moments(dV[dsmooth_idx],dI[dsmooth_idx])
# numsamp = dV[dsmooth_idx].size
# err_dm3 = np.sqrt(6*numsamp*(numsamp-1)/((numsamp-2)*(numsamp+1)*(numsamp+3)))
# print "bootstrapping..."
# moments, merrs = ADE.bootstrap(ADE.ADE_moments,[dV[dmoment_idx],dI[dmoment_idx]],10000)
# moments, merrs = ADE.bootstrap(ADE.ADE_moments,[dV[dsmooth_idx],dI[dsmooth_idx]],10000)
moments, merrs = ADE.ADE_moments(dV[dmoment_idx],dI[dmoment_idx],err=derr[dmoment_idx])
mmoments = ADE.ADE_moments(fit_mV[mmoment_idx],mI_pad[mmoment_idx])
# ADE.eplot(fit_mV[mmoment_idx],mI_pad[mmoment_idx])
# raw_input('pause')
print "moments: {}\nerrs: {}\nmmoments: {}".format(moments,merrs,mmoments)
dm3 = moments[2]
err_dm3 = merrs[2]
big_dm3 = np.append(big_dm3,dm3)
big_mm3 = np.append(big_mm3,mm3)
big_mm3_2 = np.append(big_mm3_2,mmoments[2])
big_em3 = np.append(big_em3,err_dm3)
plot_radii = np.append(plot_radii,radius)
if plotprefix:
pp.close()
pp1 = PDF(plotprefix+'_m3.pdf')
ax = plt.figure().add_subplot(111)
ax.set_title('Generated on {}'.format(datetime.now().isoformat()))
ax.set_xlabel('Radius [kpc]')
ax.set_ylabel('$\mu_3$')
# ax.plot(radii,big_dm3,'.',label='data',markersize=20,color='b')
ax.errorbar(plot_radii,big_dm3,yerr=big_em3,fmt='.',ecolor='b',ms=20,color='b')
# ax.set_ylim(-5*np.median(big_em3),5*np.median(big_em3))
# ax.set_ylim(-10,10)
ax.plot(plot_radii,big_mm3,'-',label='model',color='g')
ax.plot(plot_radii,big_mm3_2,'-',color='r',label='with same window as data')
ax.legend(loc=0,scatterpoints=1,numpoints=1)
pp1.savefig(ax.figure)
pp1.close()
plt.close('all')
return plot_radii, big_dm3, big_mm3_2, big_em3