def main():
"""Main script to prepare x-shooter observations for combination"""
from astropy.io import fits
import glob
import matplotlib.pyplot as pl
from methods import latexify
latexify()
import numpy as np
from xshoo.combine import inter_arm_cut
#Files
obj_name = 'SDSS1437-0147'
root_dir = '/Users/jselsing/Work/X-Shooter/CompositeRedQuasar/processed_data/'+obj_name
object_files = glob.glob(root_dir+'/OBJECT/*IDP*.fits')
transmission_files = glob.glob(root_dir+'/transmission*.fits')
arms = ['UVB', 'VIS', 'NIR']
wl_out = []
flux_out = []
flux_uncorr_out = []
err_out = []
start = []
end = []
for n in arms:
print('In arm: '+n)
#Read in object spectrum
obser = [k for k in object_files if n in k]
ob = fits.open(obser[0])
wl = 10.0*ob[1].data.field('WAVE')[0]
flux = ob[1].data.field('FLUX')[0]
err = ob[1].data.field('ERR')[0]
wl_tmp, flux_uncorr, err_tmp, start_tmp, end_tmp = inter_arm_cut(wl, flux, err, n, start, end)
if n== 'VIS' or n== 'NIR':
transmission = fits.open([k for k in transmission_files if n in k][0])[0].data
for j, k in enumerate(transmission):
if k <= 1e-10:
transmission[j] = 1
flux /= transmission
err /= transmission
wl, flux, err, start, end = inter_arm_cut(wl, flux, err, n, start, end)
wl_out.append(wl)
flux_out.append(flux)
err_out.append(err)
flux_uncorr_out.append(flux_uncorr)
wl_out = np.hstack(wl_out)
flux_out = np.hstack(flux_out)
err_out = np.hstack(err_out)
flux_uncorr_out = np.hstack(flux_uncorr_out)
bp_map = []
for j , (k, l) in enumerate(zip(flux_out[:-1],err_out[:-1])):
if k > 1.1 * flux_out[j-1] or k < 0:
bp_map.append(1)
elif k < 0.90 * flux_out[j-1] or k < 0:
bp_map.append(1)
else:
bp_map.append(0)
bp_map.append(1)
import json
import urllib2
query_terms = dict()
query_terms["ra"] = str(ob[0].header['RA'])+'d' #"185.1d"
query_terms["dec"] = str(ob[0].header['DEC']) #"56.78"
query_terms["radius"] = "5.0"
url = "http://api.sdss3.org/spectrumQuery?" + '&'.join(["{0}={1}".format(key, value) for key, value in query_terms.items()])
print(url)
# make call to API
response = urllib2.urlopen(url)
# read response, converting JSON format to Python list
matching_ids = json.loads(response.read())
print(json.dumps(matching_ids, indent=4))
# get the first id
spec_id = matching_ids[0]
url = "http://api.sdss3.org/spectrum?id={0}&format=json".format(spec_id)
response = urllib2.urlopen(url)
result = json.loads(response.read())
SDSS_spectrum = result[spec_id]
wl_sdss = np.array(SDSS_spectrum["wavelengths"])
flux_sdss = np.array(SDSS_spectrum["flux"])
z_sdss = (np.array(SDSS_spectrum["z"]))
z_sdss_err = ((np.array(SDSS_spectrum["z_err"])))
#Insert zeros
wl_sdss = np.concatenate([wl_sdss,np.zeros(len(wl_out) - len(wl_sdss))])
flux_sdss = np.concatenate([flux_sdss,np.zeros(len(flux_out) - len(flux_sdss))])
# Load linelist
fit_line_positions = np.genfromtxt('data/fitlinelist.txt', dtype=None)
linelist = []
for n in fit_line_positions:
linelist.append(n[1])
linelist = np.array(linelist)
from methods import wavelength_conversion
linelist = wavelength_conversion(linelist, conversion='vacuum_to_air')
#Cut out fitting region
mask = np.logical_and(wl_out > 11350, (wl_out < 11750))
wl_fit = wl_out[mask]
flux_fit = flux_out[mask]
fluxerr_fit = err_out[mask]
fluxerr_new = []
for j, (k, l) in enumerate(zip(flux_fit,fluxerr_fit)):
if k > 1.5 * flux_fit[j-2] and k > 0:
fluxerr_new.append(l*50)
elif k < 0.75 * flux_fit[j-2] and k > 0:
fluxerr_new.append(l*50)
else:
fluxerr_new.append(l)
from gen_methods import smooth
fluxerr_fit = smooth(np.array(fluxerr_new), window_len=15, window='hanning')
#Fit continuum and subtract
from methods import continuum_fit
from numpy.polynomial import chebyshev
cont, chebfit = continuum_fit(wl_fit, flux_fit, fluxerr_fit, edge_mask_len=20)
chebfitval = chebyshev.chebval(wl, chebfit)
#Define models to use
from methods import voigt,gauss
def model1(t, amp2, sig22g, sig22l, z):
tmp = voigt(t, abs(amp2), (1+z)*linelist[2], sig22g, sig22l)
return tmp
def model2(t, amp2, sig22g, z):
tmp = gauss(t, abs(amp2), (1+z)*linelist[2], sig22g)
return tmp
#Initial parameters
init_vals = [6e-12,100, z_sdss]
y_fit_guess = model2(wl_fit, *init_vals) + cont
#Fit
import scipy.optimize as op
np.random.seed(12345)
y_op = []
vals = []
for i in np.arange(10000):
print('Iteration: ', i)
resampled_spec = np.random.normal(flux_fit, abs(fluxerr_fit))
cont, chebfit = continuum_fit(wl_fit, resampled_spec, fluxerr_fit, edge_mask_len=20)
chebfitval = chebyshev.chebval(wl, chebfit)
best_vals, covar = op.curve_fit(model2, wl_fit, resampled_spec - cont, sigma=fluxerr_fit, absolute_sigma=True, p0=init_vals)
vals.append(best_vals)
up = (np.percentile(vals, 84, axis = 0)[2] - np.mean(vals, axis = 0)[2])
down = (np.percentile(vals, 16, axis = 0)[2] - np.mean(vals, axis = 0)[2])
v_bary = ob[0].header['HIERARCH ESO QC VRAD BARYCOR']
c_km = (2.99792458e8/1000.0)
print("""Curve_fit results:
Redshift = {0} + {1} - {2} (SDSS: {3} +- {4})
""".format(np.mean(vals, axis = 0)[2] + v_bary /c_km, up, down, z_sdss, z_sdss_err))
z_op = np.mean(vals, axis = 0)[2] + v_bary /c_km
#Correct for Lyman alpha forest absorption. Will only have data for objects with z > 3000 / 1216 -1 ~ 1.5
mask = (wl_out < (1 + z_op)*1216)
wave = wl_out[mask]
flux = flux_out[mask]
import continuum_mark.interactive
normalise = continuum_mark.interactive.continuum_mark(wl_out[mask], flux_out[mask], err_out[mask])
normalise.endpoint = 'n' #str(raw_input('Insert endpoint before interpolation(y/n)? '))
normalise.run()
pl.show()
cont_out = np.concatenate([normalise.continuum,flux_out[~mask]])
#Flag whether to use estimated redshift
flag = 1
from astroquery.irsa_dust import IrsaDust
import astropy.coordinates as coord
import astropy.units as u
C = coord.SkyCoord(ob[0].header['RA']*u.deg, ob[0].header['DEC']*u.deg, frame='fk5')
dust_image = IrsaDust.get_images(C, radius=2 *u.deg, image_type='ebv')[0]
ebv = np.mean(dust_image[0].data[40:42,40:42])
# Saving telluric uncorrected data to .dat file
dt = [("wl", np.float64), ("flux", np.float64), ("error", np.float64), ("bp map", np.float64),
("wl_sdss", np.float64), ("flux_sdss", np.float64) , ("flux_cont", np.float64) ]
data = np.array(zip(wl_out, flux_uncorr_out, err_out, bp_map, wl_sdss, flux_sdss, cont_out), dtype=dt)
file_name = "Telluric_uncorrected_science"
np.savetxt(root_dir+"/"+file_name+".dat", data, header="wl flux fluxerror bp_map wl_sdss flux_sdss cont")#, fmt = ['%5.1f', '%2.15E'] )
#Saving telluric corrected data to .dat file
dt = [("wl", np.float64), ("flux", np.float64), ("error", np.float64), ("bp map", np.float64),
("wl_sdss", np.float64), ("flux_sdss", np.float64) , ("flux_cont", np.float64) ]
data = np.array(zip(wl_out, flux_out, err_out, bp_map, wl_sdss, flux_sdss, cont_out), dtype=dt)
file_name = "Telluric_corrected_science"
np.savetxt(root_dir+"/"+file_name+".dat", data, header="wl flux fluxerror bp_map wl_sdss flux_sdss cont")#, fmt = ['%5.1f', '%2.15E'] )
#Saving info to .dat file
dt = [("z_op", np.float64), ("z_sdss", np.float64), ("flag", np.float64), ("ebv", np.float64)]
data = np.array(zip([z_op], [z_sdss], [flag], [ebv]), dtype=dt)
file_name = "Object_info"
np.savetxt(root_dir+"/"+file_name+".dat", data, header="z_op z_sdss flag ebv ") #, fmt = ['%5.1f', '%2.15E'] )
fluxerr_new = []
for j, (k, l) in enumerate(zip(flux_fit,fluxerr_fit)):
if k > 1.5 * flux_fit[j-2] and k > 0:
fluxerr_new.append(l*50)
elif k < 0.75 * flux_fit[j-2] and k > 0:
fluxerr_new.append(l*50)
else:
fluxerr_new.append(l)
from gen_methods import smooth
fluxerr_fit = smooth(np.array(fluxerr_new), window_len=15, window='hanning')
#Fit continuum and subtract
from methods import continuum_fit
from numpy.polynomial import chebyshev
cont, chebfit = continuum_fit(wl_fit, flux_fit, fluxerr_fit, edge_mask_len=20)
chebfitval = chebyshev.chebval(wl, chebfit)
#Define models to use
from methods import gauss
def model2(t, amp2, sig22g, z):
tmp = gauss(t, abs(amp2), (1+z)*linelist[2], sig22g)
return tmp
#Initial parameters
init_vals = [1e-15*norm,10, 0]
y_fit_guess = model2(wl_fit, *init_vals) + cont
