-
Notifications
You must be signed in to change notification settings - Fork 3
/
compute_beam.py
159 lines (130 loc) · 4.9 KB
/
compute_beam.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
import math,numpy,sys,numpy.linalg,os
import ephem
from mwapy import ephem_utils,get_observation_info
from mwapy.pb import primary_beam
from mwapy.obssched.base import schedule
import astropy.io.fits as pyfits
# open up database connection
try:
db = schedule.getdb()
except:
print "Unable to open connection to database"
sys.exit(1)
RA,Dec=69.3158945833333, -47.2523944444444
gpstime=1063400456
filename='1063400456_corr_phased.fits'
o=get_observation_info.MWA_Observation(gpstime, db=db)
mwatime=ephem_utils.MWATime(gpstime=gpstime+o.duration/2)
frequency=o.center_channel*1.28e6
delays=o.delays
RAnow,Decnow=ephem_utils.precess(RA,Dec,2000,mwatime.epoch)
HA=float(mwatime.LST)-RAnow
mwa=ephem_utils.Obs[ephem_utils.obscode['MWA']]
Az,Alt=ephem_utils.eq2horz(HA,Decnow,mwa.lat)
theta=(90-Alt)*math.pi/180
phi=Az*math.pi/180
def MWA_Jones_Analytic(theta, phi, ha, dec,
freq=100.0e6, delays=None,
zenithnorm=True):
c=2.998e8
# wavelength in meters
lam=c/freq
dip_sep=primary_beam._DIPOLE_SEPARATION
delay_int=primary_beam._DELAY_INT
dipheight=primary_beam._DIPOLE_HEIGHT
if (delays is None):
delays=0
if (isinstance(delays,float) or isinstance(delays,int)):
delays=delays*numpy.ones((16))
if (isinstance(delays,numpy.ndarray) and len(delays)==1):
delays=delays[0]*numpy.ones((16))
# direction cosines (relative to zenith) for direction az,za
projection_east=numpy.sin(theta)*numpy.sin(phi)
projection_north=numpy.sin(theta)*numpy.cos(phi)
projection_z=numpy.cos(theta)
# dipole position within the tile
dipole_north=dip_sep*numpy.array([1.5,1.5,1.5,1.5,0.5,0.5,0.5,0.5,-0.5,-0.5,-0.5,-0.5,-1.5,-1.5,-1.5,-1.5])
dipole_east=dip_sep*numpy.array([-1.5,-0.5,0.5,1.5,-1.5,-0.5,0.5,1.5,-1.5,-0.5,0.5,1.5,-1.5,-0.5,0.5,1.5])
dipole_z=dip_sep*numpy.zeros(dipole_north.shape)
# loop over dipoles
array_factor=0.0
for i in xrange(4):
for j in xrange(4):
k=4*j+i
# relative dipole phase for a source at (theta,phi)
phase=numpy.exp((1j)*2*math.pi/lam*(dipole_east[k]*projection_east + dipole_north[k]*projection_north +
dipole_z[k]*projection_z-delays[k]*c*delay_int))
array_factor+=phase/16.0
ground_plane=2*numpy.sin(2*math.pi*dipheight/lam*numpy.cos(theta))
# make sure we filter out the bottom hemisphere
ground_plane*=(theta<=math.pi/2)
# normalize to zenith
if (zenithnorm):
ground_plane/=2*numpy.sin(2*math.pi*dipheight/lam)
# this is a guess
haAntennaZenith=0
decAntennaZenith=dec-numpy.radians(mwa.lat)
sinDecAntennaZenith=numpy.sin(decAntennaZenith)
cosDecAntennaZenith=numpy.cos(decAntennaZenith)
sinDec,cosDec=numpy.sin(dec), numpy.cos(dec)
sinHa=numpy.sin(ha-haAntennaZenith)
cosHa=numpy.cos(ha-haAntennaZenith)
rot=numpy.array([cosHa,
sinDec*sinHa,
-sinDecAntennaZenith*sinHa,
cosDecAntennaZenith*cosDec+sinDecAntennaZenith*sinDec*cosHa])
gain=rot*array_factor*ground_plane
return gain
#gainXX,gainYY=primary_beam.MWA_Tile_analytic(theta, phi,
# freq=frequency*1e6,
# delays=delays)
gain=MWA_Jones_Analytic(theta, phi,
numpy.radians(HA), numpy.radians(Decnow),
freq=frequency,
delays=delays)
print gain
gain=gain.reshape((2,2))
B=numpy.linalg.inv(gain)
Bprime=B.transpose().conj()
data=pyfits.open(filename)
j=numpy.sqrt(numpy.complex(-1))
XX=data['XX-REAL'].data+j*data['XX-IMAG'].data
YY=data['YY-REAL'].data+j*data['YY-IMAG'].data
XY=data['XY-REAL'].data+j*data['XY-IMAG'].data
YX=data['YX-REAL'].data+j*data['YX-IMAG'].data
I=numpy.zeros(XX.shape)
Q=numpy.zeros(XX.shape)
U=numpy.zeros(XX.shape)
V=numpy.zeros(XX.shape)
for t in xrange(XX.shape[0]):
for f in xrange(XX.shape[1]):
#D=numpy.reshape(numpy.array([[XX[t,f].real,XY[t,f],
#XY[t,f].conj(),YY[t,f].real]]),(2,2))
D=numpy.reshape(numpy.array([[XX[t,f].real,XY[t,f],
YX[t,f],YY[t,f].real]]),(2,2))
Out=numpy.dot(numpy.dot(B,D),Bprime)
if t==5 and f==0:
out=Out
I[t,f]=0.5*(Out[1,1].real+Out[0,0].real)
Q[t,f]=0.5*(Out[1,1].real-Out[0,0].real)
U[t,f]=0.5*(Out[0,1].real+Out[1,0].real)
V[t,f]=0.5*(-Out[1,0].imag+Out[0,1].imag)
outfilename=filename.replace('.fits','_corrected.fits')
f=pyfits.PrimaryHDU(header=data[0].header)
l=[f]
l.append(pyfits.ImageHDU(data=I,
header=data[0].header))
l[-1].header['EXTNAME']='I'
l.append(pyfits.ImageHDU(data=Q,
header=data[0].header))
l[-1].header['EXTNAME']='Q'
l.append(pyfits.ImageHDU(data=U,
header=data[0].header))
l[-1].header['EXTNAME']='U'
l.append(pyfits.ImageHDU(data=V,
header=data[0].header))
l[-1].header['EXTNAME']='V'
if os.path.exists(outfilename):
os.remove(outfilename)
hl=pyfits.HDUList(l)
hl.writeto(outfilename)