def convolve_spectrum(self, cspec, exp_time, noisy=True, prng=None):
prng = parse_prng(prng)
exp_time = parse_value(exp_time, "s")
counts = cspec.flux.value * exp_time * cspec.de.value
spec = np.histogram(cspec.emid.value, self.ebins, weights=counts)[0]
conv_spec = np.zeros(self.n_ch)
pbar = tqdm(leave=True, total=self.n_e, desc="Convolving spectrum ")
if np.all(self.data["N_GRP"] == 1):
# We can do things a bit faster if there is only one group each
f_chan = ensure_numpy_array(np.nan_to_num(self.data["F_CHAN"]))
n_chan = ensure_numpy_array(np.nan_to_num(self.data["N_CHAN"]))
mat = np.nan_to_num(np.float64(self.data["MATRIX"]))
mat_size = np.minimum(n_chan, self.n_ch-f_chan)
for k in range(self.n_e):
conv_spec[f_chan[k]:f_chan[k]+n_chan[k]] += spec[k]*mat[k,:mat_size[k]]
pbar.update()
else:
# Otherwise, we have to go step-by-step
for k in range(self.n_e):
f_chan = ensure_numpy_array(np.nan_to_num(self.data["F_CHAN"][k]))
n_chan = ensure_numpy_array(np.nan_to_num(self.data["N_CHAN"][k]))
mat = np.nan_to_num(np.float64(self.data["MATRIX"][k]))
mat_size = np.minimum(n_chan, self.n_ch-f_chan)
for i, f in enumerate(f_chan):
conv_spec[f:f+n_chan[i]] += spec[k]*mat[:mat_size[i]]
pbar.update()
pbar.close()
if noisy:
return prng.poisson(lam=conv_spec)
else:
return conv_spec
def scatter_energies(self, events, prng=np.random):
"""
Scatter photon energies with the RMF and produce the
corresponding channel values.
Parameters
----------
events : dict of np.ndarrays
The energies and positions of the photons.
prng : :class:`~numpy.random.RandomState` object or :mod:`~numpy.random`, optional
A pseudo-random number generator. Typically will only be specified
if you have a reason to generate the same set of random numbers, such as for a
test. Default is the :mod:`~numpy.random` module.
"""
eidxs = np.argsort(events["energy"])
sorted_e = events["energy"][eidxs]
detectedChannels = []
# run through all photon energies and find which bin they go in
fcurr = 0
last = sorted_e.shape[0]
with ProgressBar(last) as pbar:
for (k, low), high in zip(enumerate(self.elo), self.ehi):
# weight function for probabilities from RMF
weights = np.nan_to_num(np.float64(self.data["MATRIX"][k]))
weights /= weights.sum()
# build channel number list associated to array value,
# there are groups of channels in rmfs with nonzero probabilities
trueChannel = []
f_chan = ensure_numpy_array(
np.nan_to_num(self.data["F_CHAN"][k]))
n_chan = ensure_numpy_array(
np.nan_to_num(self.data["N_CHAN"][k]))
for start, nchan in zip(f_chan, n_chan):
if nchan == 0:
trueChannel.append(start)
else:
trueChannel += list(range(start, start + nchan))
trueChannel = np.array(trueChannel)
if len(trueChannel) > 0:
e = sorted_e[fcurr:last]
nn = np.logical_and(low <= e, e < high).sum()
channelInd = prng.choice(len(weights), size=nn, p=weights)
detectedChannels.append(trueChannel[channelInd])
fcurr += nn
pbar.update(fcurr)
for key in events:
events[key] = events[key][eidxs]
events[self.header["CHANTYPE"]] = np.concatenate(detectedChannels)
return events
def _make_channels(self, k):
# build channel number list associated to array value,
# there are groups of channels in rmfs with nonzero probabilities
trueChannel = []
f_chan = ensure_numpy_array(np.nan_to_num(self.data["F_CHAN"][k]))
n_chan = ensure_numpy_array(np.nan_to_num(self.data["N_CHAN"][k]))
for start, nchan in zip(f_chan, n_chan):
if nchan == 0:
trueChannel.append(start)
else:
trueChannel += list(range(start, start + nchan))
return np.array(trueChannel)
def _make_channels(self, k):
# build channel number list associated to array value,
# there are groups of channels in rmfs with nonzero probabilities
trueChannel = []
f_chan = ensure_numpy_array(np.nan_to_num(self.data["F_CHAN"][k]))
n_chan = ensure_numpy_array(np.nan_to_num(self.data["N_CHAN"][k]))
for start, nchan in zip(f_chan, n_chan):
if nchan == 0:
trueChannel.append(start)
else:
trueChannel += list(range(start, start + nchan))
return np.array(trueChannel)
def convolve_spectrum(self, cspec, exp_time, prng=None):
prng = parse_prng(prng)
exp_time = parse_value(exp_time, "s")
counts = cspec.flux.value * exp_time * cspec.de.value
spec = np.histogram(cspec.emid.value, self.ebins, weights=counts)[0]
conv_spec = np.zeros(self.n_ch)
pbar = tqdm(leave=True, total=self.n_e, desc="Convolving spectrum ")
for k in range(self.n_e):
f_chan = ensure_numpy_array(np.nan_to_num(self.data["F_CHAN"][k]))
n_chan = ensure_numpy_array(np.nan_to_num(self.data["N_CHAN"][k]))
mat = np.nan_to_num(np.float64(self.data["MATRIX"][k]))
for f, n in zip(f_chan, n_chan):
mat_size = min(n, self.n_ch-f)
conv_spec[f:f+n] += spec[k]*mat[:mat_size]
pbar.update()
pbar.close()
return prng.poisson(lam=conv_spec)
def __init__(self, spectra, images, src_names, ra, dec, fluxes, emin, emax,
filename):
self.spectra = ensure_numpy_array(spectra)
self.images = ensure_numpy_array(images)
self.src_names = ensure_numpy_array(src_names)
self.ra = ensure_numpy_array(ra)
self.dec = ensure_numpy_array(dec)
self.fluxes = ensure_numpy_array(fluxes)
self.emin = ensure_numpy_array(emin)
self.emax = ensure_numpy_array(emax)
self.filename = filename
self.num_sources = self.spectra.size
# timing not yet supported
self.timing = np.array(["NULL"] * self.num_sources)
def make_background(exp_time,
instrument,
sky_center,
foreground=True,
ptsrc_bkgnd=True,
instr_bkgnd=True,
no_dither=False,
dither_params=None,
roll_angle=0.0,
subpixel_res=False,
input_sources=None,
absorb_model="wabs",
nH=0.05,
aimpt_shift=None,
prng=None):
"""
Make background events.
Parameters
----------
exp_time : float, (value, unit) tuple, or :class:`~astropy.units.Quantity`
The exposure time to use, in seconds.
instrument : string
The name of the instrument to use, which picks an instrument
specification from the instrument registry.
sky_center : array, tuple, or list
The center RA, Dec coordinates of the observation, in degrees.
foreground : boolean, optional
Whether or not to include the Galactic foreground. Default: True
instr_bkgnd : boolean, optional
Whether or not to include the instrumental background. Default: True
no_dither : boolean, optional
If True, turn off dithering entirely. Default: False
dither_params : array-like of floats, optional
The parameters to use to control the size and period of the dither
pattern. The first two numbers are the dither amplitude in x and y
detector coordinates in arcseconds, and the second two numbers are
the dither period in x and y detector coordinates in seconds.
Default: [8.0, 8.0, 1000.0, 707.0].
ptsrc_bkgnd : boolean, optional
Whether or not to include the point-source background. Default: True
Default: 0.05
roll_angle : float, (value, unit) tuple, or :class:`~astropy.units.Quantity`, optional
The roll angle of the observation in degrees. Default: 0.0
subpixel_res: boolean, optional
If True, event positions are not randomized within the pixels
within which they are detected. Default: False
input_sources : string, optional
If set to a filename, input the point source positions, fluxes,
and spectral indices from an ASCII table instead of generating
them. Default: None
absorb_model : string, optional
The absorption model to use, "wabs" or "tbabs". Default: "wabs"
nH : float, optional
The hydrogen column in units of 10**22 atoms/cm**2.
Default: 0.05
aimpt_shift : array-like, optional
A two-float array-like object which shifts the aimpoint on the
detector from the nominal position. Units are in arcseconds.
Default: None, which results in no shift from the nominal aimpoint.
prng : :class:`~numpy.random.RandomState` object, integer, or None
A pseudo-random number generator. Typically will only
be specified if you have a reason to generate the same
set of random numbers, such as for a test. Default is None,
which sets the seed based on the system time.
"""
from soxs.background import make_instrument_background, \
make_foreground, make_ptsrc_background
prng = parse_prng(prng)
exp_time = parse_value(exp_time, "s")
roll_angle = parse_value(roll_angle, "deg")
try:
instrument_spec = instrument_registry[instrument]
except KeyError:
raise KeyError(f"Instrument {instrument} is not in the "
f"instrument registry!")
if not instrument_spec["imaging"]:
raise RuntimeError(f"Instrument '{instrument_spec['name']}' is not "
f"designed for imaging observations!")
fov = instrument_spec["fov"]
input_events = defaultdict(list)
arf_file = get_data_file(instrument_spec["arf"])
arf = AuxiliaryResponseFile(arf_file)
rmf_file = get_data_file(instrument_spec["rmf"])
rmf = RedistributionMatrixFile(rmf_file)
if ptsrc_bkgnd:
mylog.info("Adding in point-source background.")
ptsrc_events = make_ptsrc_background(exp_time,
fov,
sky_center,
area=1.2 * arf.max_area,
input_sources=input_sources,
absorb_model=absorb_model,
nH=nH,
prng=prng)
for key in ["ra", "dec", "energy"]:
input_events[key].append(ptsrc_events[key])
input_events["flux"].append(ptsrc_events["flux"])
input_events["emin"].append(ptsrc_events["energy"].min())
input_events["emax"].append(ptsrc_events["energy"].max())
input_events["src_names"].append("ptsrc_bkgnd")
events, event_params = generate_events(input_events,
exp_time,
instrument,
sky_center,
no_dither=no_dither,
dither_params=dither_params,
roll_angle=roll_angle,
subpixel_res=subpixel_res,
aimpt_shift=aimpt_shift,
prng=prng)
mylog.info(f"Generated {events['energy'].size} photons from "
f"the point-source background.")
else:
nx = instrument_spec["num_pixels"]
plate_scale = instrument_spec["fov"] / nx / 60.0
plate_scale_arcsec = plate_scale * 3600.0
if aimpt_shift is None:
aimpt_shift = np.zeros(2)
aimpt_shift = ensure_numpy_array(aimpt_shift).astype('float64')
aimpt_shift /= plate_scale_arcsec
events = defaultdict(list)
if not instrument_spec["dither"]:
dither_on = False
else:
dither_on = not no_dither
if dither_params is None:
dither_params = [8.0, 8.0, 1000.0, 707.0]
dither_dict = {
"x_amp": dither_params[0],
"y_amp": dither_params[1],
"x_period": dither_params[2],
"y_period": dither_params[3],
"dither_on": dither_on,
"plate_scale": instrument_spec["fov"] / nx * 60.0
}
event_params = {
"exposure_time": exp_time,
"fov": instrument_spec["fov"],
"num_pixels": nx,
"pix_center": np.array([0.5 * (2 * nx + 1)] * 2),
"channel_type": rmf.header["CHANTYPE"],
"sky_center": sky_center,
"dither_params": dither_dict,
"plate_scale": plate_scale,
"chan_lim": [rmf.cmin, rmf.cmax],
"rmf": rmf_file,
"arf": arf_file,
"telescope": rmf.header["TELESCOP"],
"instrument": instrument_spec['name'],
"mission": rmf.header.get("MISSION", ""),
"nchan": rmf.n_ch,
"roll_angle": roll_angle,
"aimpt_coords": instrument_spec["aimpt_coords"],
"aimpt_shift": aimpt_shift
}
if "chips" not in event_params:
event_params["chips"] = instrument_spec["chips"]
if foreground:
mylog.info("Adding in astrophysical foreground.")
bkg_events = make_foreground(event_params, arf, rmf, prng=prng)
for key in bkg_events:
events[key] = np.concatenate([events[key], bkg_events[key]])
if instr_bkgnd and instrument_spec["bkgnd"] is not None:
mylog.info("Adding in instrumental background.")
bkg_events = make_instrument_background(instrument_spec,
event_params,
rmf,
prng=prng)
for key in bkg_events:
events[key] = np.concatenate([events[key], bkg_events[key]])
return events, event_params
def generate_events(source,
exp_time,
instrument,
sky_center,
no_dither=False,
dither_params=None,
roll_angle=0.0,
subpixel_res=False,
aimpt_shift=None,
prng=None):
"""
Take unconvolved events and convolve them with instrumental responses. This
function does the following:
1. Determines which events are observed using the ARF
2. Pixelizes the events, applying PSF effects and dithering
3. Determines energy channels using the RMF
This function is not meant to be called by the end-user but is used by
the :func:`~soxs.instrument.instrument_simulator` function.
Parameters
----------
input_events : string, dict, or None
The unconvolved events to be used as input. Can be one of the
following:
1. The name of a SIMPUT catalog file.
2. A Python dictionary containing the following items:
"ra": A NumPy array of right ascension values in degrees.
"dec": A NumPy array of declination values in degrees.
"energy": A NumPy array of energy values in keV.
"flux": The flux of the entire source, in units of erg/cm**2/s.
out_file : string
The name of the event file to be written.
exp_time : float, (value, unit) tuple, or :class:`~astropy.units.Quantity`
The exposure time to use, in seconds.
instrument : string
The name of the instrument to use, which picks an instrument
specification from the instrument registry.
sky_center : array, tuple, or list
The center RA, Dec coordinates of the observation, in degrees.
no_dither : boolean, optional
If True, turn off dithering entirely. Default: False
dither_params : array-like of floats, optional
The parameters to use to control the size and period of the dither
pattern. The first two numbers are the dither amplitude in x and y
detector coordinates in arcseconds, and the second two numbers are
the dither period in x and y detector coordinates in seconds.
Default: [8.0, 8.0, 1000.0, 707.0].
roll_angle : float, (value, unit) tuple, or :class:`~astropy.units.Quantity`, optional
The roll angle of the observation in degrees. Default: 0.0
subpixel_res : boolean, optional
If True, event positions are not randomized within the pixels
within which they are detected. Default: False
aimpt_shift : array-like, optional
A two-float array-like object which shifts the aimpoint on the
detector from the nominal position. Units are in arcseconds.
Default: None, which results in no shift from the nominal aimpoint.
prng : :class:`~numpy.random.RandomState` object, integer, or None
A pseudo-random number generator. Typically will only
be specified if you have a reason to generate the same
set of random numbers, such as for a test. Default is None,
which sets the seed based on the system time.
"""
exp_time = parse_value(exp_time, "s")
roll_angle = parse_value(roll_angle, "deg")
prng = parse_prng(prng)
if source is None:
source_list = []
elif isinstance(source, dict):
parameters = {}
for key in ["flux", "emin", "emax", "src_names"]:
parameters[key] = source[key]
source_list = []
for i in range(len(parameters["flux"])):
phlist = SimputPhotonList(source["ra"][i], source["dec"][i],
source["energy"][i],
parameters['flux'][i],
parameters['src_names'][i])
source_list.append(phlist)
elif isinstance(source, str):
# Assume this is a SIMPUT catalog
source_list, parameters = read_simput_catalog(source)
try:
instrument_spec = instrument_registry[instrument]
except KeyError:
raise KeyError(
f"Instrument {instrument} is not in the instrument registry!")
if not instrument_spec["imaging"]:
raise RuntimeError(f"Instrument '{instrument_spec['name']}' is not "
f"designed for imaging observations!")
arf_file = get_data_file(instrument_spec["arf"])
rmf_file = get_data_file(instrument_spec["rmf"])
arf = AuxiliaryResponseFile(arf_file)
rmf = RedistributionMatrixFile(rmf_file)
nx = instrument_spec["num_pixels"]
plate_scale = instrument_spec["fov"] / nx / 60. # arcmin to deg
plate_scale_arcsec = plate_scale * 3600.0
if aimpt_shift is None:
aimpt_shift = np.zeros(2)
aimpt_shift = ensure_numpy_array(aimpt_shift).astype('float64')
aimpt_shift /= plate_scale_arcsec
if not instrument_spec["dither"]:
dither_on = False
else:
dither_on = not no_dither
if dither_params is None:
dither_params = [8.0, 8.0, 1000.0, 707.0]
dither_dict = {
"x_amp": dither_params[0],
"y_amp": dither_params[1],
"x_period": dither_params[2],
"y_period": dither_params[3],
"dither_on": dither_on,
"plate_scale": plate_scale_arcsec
}
event_params = {
"exposure_time": exp_time,
"arf": arf.filename,
"sky_center": sky_center,
"pix_center": np.array([0.5 * (2 * nx + 1)] * 2),
"num_pixels": nx,
"plate_scale": plate_scale,
"rmf": rmf.filename,
"channel_type": rmf.chan_type,
"telescope": rmf.header["TELESCOP"],
"instrument": instrument_spec['name'],
"mission": rmf.header.get("MISSION", ""),
"nchan": rmf.n_ch,
"roll_angle": roll_angle,
"fov": instrument_spec["fov"],
"chan_lim": [rmf.cmin, rmf.cmax],
"chips": instrument_spec["chips"],
"dither_params": dither_dict,
"aimpt_coords": instrument_spec["aimpt_coords"],
"aimpt_shift": aimpt_shift
}
# Set up WCS
w = pywcs.WCS(naxis=2)
w.wcs.crval = event_params["sky_center"]
w.wcs.crpix = event_params["pix_center"]
w.wcs.cdelt = [-plate_scale, plate_scale]
w.wcs.ctype = ["RA---TAN", "DEC--TAN"]
w.wcs.cunit = ["deg"] * 2
# Determine rotation matrix
rot_mat = get_rot_mat(roll_angle)
# Set up PSF
psf_type = instrument_spec["psf"][0]
psf_class = psf_model_registry[psf_type]
psf = psf_class(instrument_spec, prng=prng)
all_events = defaultdict(list)
for i, src in enumerate(source_list):
mylog.info(
f"Detecting events from source {parameters['src_names'][i]}")
# Step 1: Use ARF to determine which photons are observed
mylog.info(f"Applying energy-dependent effective area from "
f"{os.path.split(arf.filename)[-1]}.")
refband = [parameters["emin"][i], parameters["emax"][i]]
if src.src_type == "phlist":
events = arf.detect_events_phlist(src.events.copy(),
exp_time,
parameters["flux"][i],
refband,
prng=prng)
elif src.src_type.endswith("spectrum"):
events = arf.detect_events_spec(src, exp_time, refband, prng=prng)
n_evt = events["energy"].size
if n_evt == 0:
mylog.warning("No events were observed for this source!!!")
else:
# Step 2: Assign pixel coordinates to events. Apply dithering and
# PSF. Clip events that don't fall within the detection region.
mylog.info("Pixeling events.")
# Convert RA, Dec to pixel coordinates
xpix, ypix = w.wcs_world2pix(events["ra"], events["dec"], 1)
xpix -= event_params["pix_center"][0]
ypix -= event_params["pix_center"][1]
events.pop("ra")
events.pop("dec")
n_evt = xpix.size
# Rotate physical coordinates to detector coordinates
det = np.dot(rot_mat, np.array([xpix, ypix]))
detx = det[0, :] + event_params["aimpt_coords"][0] + aimpt_shift[0]
dety = det[1, :] + event_params["aimpt_coords"][1] + aimpt_shift[1]
# Add times to events
events['time'] = prng.uniform(size=n_evt,
low=0.0,
high=event_params["exposure_time"])
# Apply dithering
x_offset, y_offset = perform_dither(events["time"], dither_dict)
detx -= x_offset
dety -= y_offset
# PSF scattering of detector coordinates
mylog.info(f"Scattering events with a {psf}-based PSF.")
detx, dety = psf.scatter(detx, dety, events["energy"])
# Convert detector coordinates to chip coordinates.
# Throw out events that don't fall on any chip.
cx = np.trunc(detx) + 0.5 * np.sign(detx)
cy = np.trunc(dety) + 0.5 * np.sign(dety)
events["chip_id"] = -np.ones(n_evt, dtype='int')
for i, chip in enumerate(event_params["chips"]):
rtype = chip[0]
args = chip[1:]
r, _ = create_region(rtype, args, 0.0, 0.0)
inside = r.contains(PixCoord(cx, cy))
events["chip_id"][inside] = i
keep = events["chip_id"] > -1
mylog.info(f"{n_evt-keep.sum()} events were rejected because "
f"they do not fall on any CCD.")
n_evt = keep.sum()
if n_evt == 0:
mylog.warning("No events are within the field "
"of view for this source!!!")
else:
# Keep only those events which fall on a chip
for key in events:
events[key] = events[key][keep]
# Convert chip coordinates back to detector coordinates,
# unless the user has specified that they want subpixel
# resolution
if subpixel_res:
events["detx"] = detx[keep]
events["dety"] = dety[keep]
else:
events["detx"] = cx[keep] + \
prng.uniform(low=-0.5, high=0.5, size=n_evt)
events["dety"] = cy[keep] + \
prng.uniform(low=-0.5, high=0.5, size=n_evt)
# Convert detector coordinates back to pixel coordinates by
# adding the dither offsets back in and applying the rotation
# matrix again
det = np.array([
events["detx"] + x_offset[keep] -
event_params["aimpt_coords"][0] - aimpt_shift[0],
events["dety"] + y_offset[keep] -
event_params["aimpt_coords"][1] - aimpt_shift[1]
])
pix = np.dot(rot_mat.T, det)
events["xpix"] = pix[0, :] + event_params['pix_center'][0]
events["ypix"] = pix[1, :] + event_params['pix_center'][1]
if n_evt > 0:
for key in events:
all_events[key] = np.concatenate(
[all_events[key], events[key]])
if len(all_events["energy"]) == 0:
mylog.warning("No events from any of the sources in "
"the catalog were detected!")
for key in [
"xpix", "ypix", "detx", "dety", "time", "chip_id",
event_params["channel_type"]
]:
all_events[key] = np.array([])
else:
# Step 4: Scatter energies with RMF
mylog.info(f"Scattering energies with "
f"RMF {os.path.split(rmf.filename)[-1]}.")
all_events = rmf.scatter_energies(all_events, prng=prng)
return all_events, event_params