Plotting with Joe

Joe Hennawi and I worked on trying to figure out where exactly we are going wrong with the likelihood method. He suggested making a two-dimensional histogram of the likelihoods binned in color-color space. This would allow us to see where in color-color space were are getting objects which have high likelihoods and will help us debug if there is a population of objects which are being falsely selected.

It is quite easy to make the 2D histogram, but plotting it has proved to be an issue. I've spent most of the day trying to do this. Here is how I create the histogram:

qsos = testqsos

b_u = 1.4
b_g = 0.9
b_r = 1.2
b_i = 1.8
b_z = 7.4

z_temp = qsos.Z_TOT
fu1 = qsos.PSFFLUX[0]
fg1 = qsos.PSFFLUX[1]
fr1 = qsos.PSFFLUX[2]
fi1 = qsos.PSFFLUX[3]
fz1 = qsos.PSFFLUX[4]

ivar_u1 = qsos.PSFFLUX_IVAR[0]
ivar_g1 = qsos.PSFFLUX_IVAR[1]
ivar_r1 = qsos.PSFFLUX_IVAR[2]
ivar_i1 = qsos.PSFFLUX_IVAR[3]
ivar_z1 = qsos.PSFFLUX_IVAR[4]


u1 = sdss_flux2mags(fu1, b_u)
g1 = sdss_flux2mags(fg1, b_g)
r1 = sdss_flux2mags(fr1, b_r)
i1 = sdss_flux2mags(fi1, b_i)
z1 = sdss_flux2mags(fz1, b_z)

sigu = sdss_ivar2magerr(ivar_u1, fu1, b_u)
sigg = sdss_ivar2magerr(ivar_g1, fg1, b_g)
sigr = sdss_ivar2magerr(ivar_r1, fr1, b_r)
sigi = sdss_ivar2magerr(ivar_i1, fi1, b_i)
sigz = sdss_ivar2magerr(ivar_z1, fz1, b_z)

ug = u1-g1
gr = g1-r1
ri = r1-i1
iz = i1-z1

icut = where(i1 GT 19.5)

nx = 50
ny = 50
imageQSO = fltarr(nx,ny)
imageAll = fltarr(nx,ny)
imageRatio = fltarr(nx,ny)

likeQSO = qsoqlike[icut]
likeAll = qsoslike[icut]
likeRatio = likeQSO/likeAll

qsoug = ug[icut]
qsogr = gr[icut]

populate_image, imageQSO, qsoug, qsogr, weight = likeQSO
populate_image, imageALL, qsoug, qsogr, weight = likeAll
populate_image, imageRatio, qsoug, qsogr, weight = likeRatio

Now I need code to plot this histogram in color-color space with hotter colors in bins with higher likelihoods and cooler colors in bins with smaller likelihoods. Anyone have code that does this in IDL or python? Joe has some, but I'm having a hard time getting it to work.