forked from amaurea/tenki
-
Notifications
You must be signed in to change notification settings - Fork 0
/
simtod.py
157 lines (148 loc) · 6.46 KB
/
simtod.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
import numpy as np, argparse, enlib.scan, os
from enlib import enmap, utils, config, scansim, log, powspec, fft, bunch
from enact import data, filedb, nmat_measure
config.default("verbosity", 1, "Verbosity for output. Higher means more verbose. 0 outputs only errors etc. 1 outputs INFO-level and 2 outputs DEBUG-level messages.")
parser = config.ArgumentParser(os.environ["HOME"] + "/.enkirc")
parser.add_argument("odir")
parser.add_argument("--area", type=str)
parser.add_argument("--bore", type=str, default="grid:2:0.2:0.8")
parser.add_argument("--dets", type=str, default="scattered:3:3:2.0")
parser.add_argument("--signal",type=str, default="ptsrc:100:1e3:-3")
parser.add_argument("--noise", type=str, default="1/f:20:2:0.5")
parser.add_argument("--seed", type=int, default=1)
parser.add_argument("--measure", type=float, default=None)
parser.add_argument("--real", type=str, default=None)
args = parser.parse_args()
log_level = log.verbosity2level(config.get("verbosity"))
L = log.init(level=log_level)
utils.mkdir(args.odir)
if args.area:
area = enmap.read_map(args.area)
if area.ndim == 2: area = area[None]
else:
shape, wcs = enmap.geometry(pos=np.array([[-1,-1],[1,1]])*np.pi/180, shape=(600,600), pre=(3,), proj="car", ref=[0,0])
area = enmap.zeros(shape, wcs)
def get_scans(area, signal, bore, dets, noise, seed=0, real=None, noise_override=None):
scans = []
# Get real scan information if necessary
L.debug("real")
if real:
real_scans = []
filedb.init()
db = filedb.data
ids = fileb.scans[real].ids
for id in ids:
try:
real_scans.append(data.ACTScan(db[id]))
except errors.DataMissing as e:
L.debug("Skipped %s (%s)" % (id, e.message))
# Dets
L.debug("dets")
sim_dets = []
toks = dets.split(":")
if toks[0] == "scattered":
ngroup, nper, rad = int(toks[1]), int(toks[2]), float(toks[3])
sim_dets = [scansim.dets_scattered(ngroup, nper,rad=rad*np.pi/180/60)]
margin = rad*np.pi/180/60
elif toks[0] == "real":
ndet = int(toks[1])
dslice = slice(0,ndet) if ndet > 0 else slice(None)
sim_dets = [bunch.Bunch(comps=s.comps[dslice], offsets=s.offsets[dslice]) for s in real_scans]
margin = np.max([np.sum(s.offsets**2,1)**0.5 for s in sim_dets])
else: raise ValueError
# Boresight. Determines our number of scans
L.debug("bore")
sim_bore = []
toks = bore.split(":")
if toks[0] == "grid":
nscan, density, short = int(toks[1]), float(toks[2]), float(toks[3])
for i in range(nscan):
tbox = shorten(area.box(),i%2,short)
sim_bore.append(scansim.scan_grid(tbox, density*np.pi/180/60, dir=i, margin=margin))
elif toks[0] == "real":
sim_bore = [bunch.Bunch(boresight=s.boresight, sys=s.sys, site=s.site, mjd0=s.mjd0) for s in real_scans]
else: raise ValueError
nsim = len(sim_bore)
# Make one det info per scan
sim_dets = sim_dets*(nsim/len(sim_dets))+sim_dets[:nsim%len(sim_dets)]
# Noise
L.debug("noise")
sim_nmat = []
toks = noise.split(":")
nonoise = False
if toks[0] == "1/f":
sigma, alpha, fknee = [float(v) for v in toks[1:4]]
nonoise = sigma < 0
for i in range(nsim):
sim_nmat.append(scansim.oneoverf_noise(sim_dets[i].comps.shape[0], sim_bore[i].boresight.shape[0], sigma=np.abs(sigma), alpha=alpha, fknee=fknee))
elif toks[0] == "detcorr":
sigma, alpha, fknee = [float(v) for v in toks[1:4]]
nonoise = sigma < 0
for i in range(nsim):
sim_nmat.append(scansim.oneoverf_detcorr_noise(sim_dets[i].comps.shape[0], sim_bore[i].boresight.shape[0], sigma=np.abs(sigma), alpha=alpha, fknee=fknee))
elif toks[0] == "real":
scale = 1.0 if len(toks) < 2 else float(toks[1])
for i,s in enumerate(real_scans):
ndet = len(sim_dets[i].offsets)
nmat = s.noise[:ndet]*scale**-2
sim_nmat.append(nmat)
else: raise ValueError
noise_scale = not nonoise if noise_override is None else noise_override
sim_nmat = sim_nmat*(nsim/len(sim_nmat))+sim_nmat[:nsim%len(sim_nmat)]
# Signal
L.debug("signal")
toks = signal.split(":")
if toks[0] == "ptsrc":
# This one always operates in the same coordinates as
nsrc, amp, fwhm = int(toks[1]), float(toks[2]), float(toks[3])
np.random.seed(seed)
sim_srcs = scansim.rand_srcs(area.box(), nsrc, amp, abs(fwhm)*np.pi/180/60, rand_fwhm=fwhm<0)
for i in range(nsim):
scans.append(scansim.SimSrcs(sim_bore[i], sim_dets[i], sim_srcs, sim_nmat[i], seed=seed+i, noise_scale=noise_scale))
elif toks[0] == "vsrc":
# Create a single variable source
ra, dec, fwhm = float(toks[1])*np.pi/180, float(toks[2])*np.pi/180, float(toks[3])*np.pi/180/60
amps = [float(t) for t in toks[4].split(",")]
for i in range(nsim):
sim_srcs = bunch.Bunch(pos=np.array([[dec,ra]]),amps=np.array([[amps[i],0,0,0]]), beam=np.array([fwhm/(8*np.log(2)**0.5)]))
scans.append(scansim.SimSrcs(sim_bore[i], sim_dets[i], sim_srcs, sim_nmat[i], seed=seed+i, noise_scale=noise_scale, nsigma=20))
elif toks[0] == "cmb":
np.random.seed(seed)
ps = powspec.read_spectrum(toks[1])
sim_map = enmap.rand_map(area.shape, area.wcs, ps)
for i in range(nsim):
scans.append(scansim.SimMap(sim_bore[i], sim_dets[i], sim_map, sim_nmat[i], seed=seed+i, noise_scale=noise_scale))
else: raise ValueError
return scans
def shorten(box, i, s=0.5):
box = np.array(box)
m = (box[1]+box[0])/2
w = (box[1]-box[0])/2*s
j = i%2
box[:,j] = [m[j]-w[j],m[j]+w[j]]
return box
def get_model(s, area):
pos = area.posmap().reshape(2,-1)[::-1].T
model = np.rollaxis(s.get_model(pos),-1).reshape(-1,area.shape[1],area.shape[2])
return enmap.ndmap(model, area.wcs)[:area.shape[0]]
if args.measure is None:
scans = get_scans(area, args.signal, args.bore, args.dets, args.noise, seed=args.seed, real=args.real)
else:
# Build noise model the same way we do for the real data, i.e. based on
# measuring data itself. But do that based on a version with more noise
# than the real one, to simulate realistic S/N ratios without needing
# too many samples
scans = get_scans(area, args.signal, args.bore, args.dets, args.noise, seed=args.seed, real=args.real, noise_override=args.measure)
nmats = []
for scan in scans:
ft = fft.rfft(scan.get_samples()) * scan.nsamp**-0.5
nmats.append(nmat_measure.detvecs_jon(ft, 400.0, shared=True))
scans = get_scans(area, args.signal, args.bore, args.dets, args.noise, seed=args.seed, real=args.real)
for scan,nmat in zip(scans,nmats):
scan.noise = nmat
enmap.write_map(args.odir + "/area.fits", area)
model = get_model(scans[0], area)
enmap.write_map(args.odir + "/model.fits", model)
for i, scan in enumerate(scans):
L.info("scan %2d/%d" % (i+1,len(scans)))
enlib.scan.write_scan(args.odir + "/scan%03d.hdf" % i, scan)