import mosaic
import pyfits
pos = mosaic.read_positions()
tables = dict(
Hgamma = mosaic.read_fits_table("Hgamma_mosaic_sb.fits"),
Hdelta = mosaic.read_fits_table("Hdelta_mosaic_sb.fits")
)
# Write a FITS file for each data column of each emission line
for emline, table in tables.items():
for name in table.names:
im = mosaic.interpolate_image(pos.x, pos.y, table[name], delta=0.5)
pyfits.writeto(
"%s-%s.fits" % (emline, name),
im[:,::-1],
clobber=True)
# And do the flux ratio of the two lines
ratio = tables["Hdelta"].flux / tables["Hgamma"].flux
im = mosaic.interpolate_image(pos.x, pos.y, ratio, delta=0.5)
pyfits.writeto("Hd-Hg-ratio.fits", im[:,::-1], clobber=True)
def compare(lineid, specid, column='flux', posid="positions_mosaic"):
fieldpairs = mosaic.find_overlaps(posid=posid)
fitsfilename = "%s_mosaic_%s.fits" % (lineid, specid)
table = mosaic.read_fits_table(fitsfilename)
# initialise a table to save the results
outtable = dict(field1=list(), field2=list(),
s_mean=list(), s_std=list(), N=list(),
rat_med=list(), rat_q25=list(), rat_q75=list())
for fieldpair, indices in fieldpairs.items():
field, otherfield = fieldpair
values = table[column][[i-1 for i in indices.keys()]]
otherindices = list()
seps = list()
for neighbours in indices.values():
# first attempt - just use the nearest neighbour
sep, otherindex = neighbours[0]
otherindices.append(otherindex)
seps.append(sep)
othervalues = table[column][[i-1 for i in otherindices]]
seps = np.array(seps)
# plot the data
try:
datamax = 1.1*max(values.max(), othervalues.max())
except ValueError:
# something wrong with this pair: skip it
continue
# The color of the points represents the separations
plt.scatter(x=values, y=othervalues, c=seps,
marker="o", vmin=0.0, vmax=2.68, alpha=0.6)
plt.axis([0.0, datamax, 0.0, datamax])
cb = plt.colorbar()
cb.set_label("aperture\nseparation")
# plt.axis('scaled')
plt.xlabel(field)
plt.ylabel(otherfield)
# plot line y =x
x = np.linspace(0.0, datamax)
plt.plot(x, x, 'r--')
# Median gradient
medgrad = np.median(othervalues/values)
quart25 = scipy.stats.scoreatpercentile(othervalues/values, 25)
quart75 = scipy.stats.scoreatpercentile(othervalues/values, 75)
dyplus = quart75 - medgrad
dyminus = medgrad - quart25
y = medgrad*x
plt.plot(x, y, 'g')
# Add info to title
plt.title("%s %s %s\n N = %i, median(%s/%s) = %.2f + %.2f - %.2f\n sep = %.2f +/- %.2f"
% (lineid, specid, column, len(values),
otherfield, field, medgrad, dyplus, dyminus,
seps.mean(), seps.std()),
fontsize="small")
plt.savefig("%s-%s-compare-%s-%s-%s.png"
% (lineid, specid, column, field, otherfield))
plt.clf()
# save the data to the output table
outtable["field1"].append(field)
outtable["field2"].append(otherfield)
outtable["s_mean"].append(seps.mean())
outtable["s_std"].append(seps.std())
outtable["N"].append(len(values))
outtable["rat_med"].append(medgrad)
outtable["rat_q25"].append(quart25)
outtable["rat_q75"].append(quart75)
# Write output table to file
fmt = "%.3f"
fmts = dict(s_mean=fmt, s_std=fmt,
rat_med=fmt, rat_q25=fmt, rat_q75=fmt)
asciitable.write(outtable,
"%s-%s-compare-%s.dat" % (lineid, specid, column),
delimiter="\t", formats=fmts)
