def _check_transform(coords):
# Coords is a shape.Subinterval
wcsprm = Wcsprm()
coord_array = np.array([[coords.lower, coords.upper]])
sky_transform = wcsprm.p2s(coord_array, ORIGIN)
pix_transform = wcsprm.s2p(sky_transform['world'], ORIGIN)
transformed_coords = pix_transform['pixcrd']
if not (transformed_coords[0][0] == coords.lower
and transformed_coords[0][1] == coords.upper):
raise ValueError(
"Could not transform coordinates pixel to sky, sky to pixel")
def _check_transform(coords):
# Coords is a shape.Subinterval
wcsprm = Wcsprm()
coord_array = np.array([[coords.lower, coords.upper]])
sky_transform = wcsprm.p2s(coord_array, ORIGIN)
wcsprm.s2p(sky_transform['world'], ORIGIN)
def _check_transform(lower, upper):
wcsprm = Wcsprm()
coord_array = np.array([[lower, upper]])
sky_transform = wcsprm.p2s(coord_array, ORIGIN)
wcsprm.s2p(sky_transform['world'], ORIGIN)
def _check_transform(lower, upper):
wcsprm = Wcsprm()
coord_array = np.array([[lower, upper]])
sky_transform = wcsprm.p2s(coord_array, ORIGIN)
wcsprm.s2p(sky_transform['world'], ORIGIN)
def main():
"""Perform photometry for the given file."""
print("Program version: 0.2")
StartTime = datetime.now()
args = parseArguments()
if (args.ignore_warnings):
warnings.simplefilter('ignore', UserWarning)
fits_image_filenames = args.input
#print(fits_image_filenames)
#for directories search for appropriate fits files
if (os.path.isdir(fits_image_filenames[0])):
print(
"detected a directory. Will search for fits files in it that already have astrometry calibration and therefor contain _astro"
)
path = fits_image_filenames[0]
fits_image_filenames = []
for file in os.listdir(path):
if file.endswith(".fits") and "_astro" in file:
fits_image_filenames.append(path + "/" + file)
print(fits_image_filenames)
for fits_image_filename in fits_image_filenames:
print("")
print(">>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>>")
print("> Photometry for {} ".format(fits_image_filename))
with fits.open(fits_image_filename) as hdul:
#print(hdul.info())
if (args.verbose):
print(
"if image is not at first position in the fits file the program will break later on"
)
#print(hdul[0].header)
hdu = hdul[0]
hdr = hdu.header
image_or = hdul[0].data.astype(float)
image = image_or - np.median(image_or)
wcsprm = Wcsprm(hdr.tostring().encode(
'utf-8')) #everything else gave me errors with python 3
#tranlating aperture into pixel:
on_sky = wcsprm.p2s([[0, 0], [1, 1]], 0)["world"]
px_scale = np.sqrt((on_sky[0, 0] - on_sky[1, 0])**2 +
(on_sky[0, 1] - on_sky[1, 1])**2)
px_scale = px_scale * 60 * 60 #in arcsec
aperture = args.aperture / px_scale #aperture n pixel
print("aperture")
observation = find_sources(image, aperture)
#print(observation)
positions = (observation['xcenter'], observation['ycenter'])
apertures = CircularAperture(positions, r=4.)
#get rough coordinates
coord = SkyCoord(wcsprm.crval[0],
wcsprm.crval[1],
unit=(u.deg, u.deg),
frame="icrs")
#put in nice wrapper! with repeated tries and maybe try synchron!
print(">Dowloading catalog data")
#WCS.calc_footprint(header=None, undistort=True, axes=None, center=True)
radius = u.Quantity(5, u.arcmin) #should be enough for all images
catalog_data, band_name, catalog_name, mag_sys = query.get_photometry_data(
coord, radius, args.band, args.catalog)
#throwing out blended sources (should be improved, TODO)
apertures_catalog = CircularAperture(wcsprm.s2p(
catalog_data[["ra", "dec"]], 1)['pixcrd'],
r=5.)
obs_matched, cat_matched, distances = register.find_matches_keep_catalog_info(
observation, catalog_data, wcsprm, threshold=3)
print("Found {} matches".format(obs_matched.shape[0]))
MAG_CALC = True
if (obs_matched.shape[0] == 0):
MAG_CALC = False
obs_matched["aperture_sum"]
cat_matched[band_name]
#mag = -2.5 log10(cts) + ZP
ZP = -1 * (-2.5 * np.log10(obs_matched["aperture_sum"].values) -
cat_matched[band_name].values)
#FIGURE OUT LINEAR RANGE TOTO
ZP_median = np.median(ZP[~np.isnan(ZP)])
new_magnitudes = -2.5 * np.log10(
observation["aperture_sum"].values) + ZP_median
ra = args.ra
dec = args.dec
if (args.name):
targets = pd.read_csv("targets.csv")
if ((targets["NAME"] == args.name).sum()):
ra = targets.loc[targets["NAME"] == args.name, "RA"].values[0]
dec = targets.loc[targets["NAME"] == args.name,
"DEC"].values[0]
print("For {} the following coordinates where found".format(
args.name))
print("ra: {} dec: {}".format(ra, dec))
else:
print("{} not found".format(args.name))
if (ra and dec):
c_unknown = SkyCoord(ra, dec, unit=(u.deg, u.deg), frame="icrs")
#print(observation[["xcenter", "ycenter"]].values)
coordinats_obs = wcsprm.p2s(
observation[["xcenter", "ycenter"]].values, 1)["world"]
#print(coordinats_obs[17])
c_obs = SkyCoord(coordinats_obs[:, 0],
coordinats_obs[:, 1],
unit=(u.deg, u.deg),
frame="icrs")
#print(coordinats_obs)
from astropy.coordinates import match_coordinates_sky
idx, d2d, d3d = match_coordinates_sky(c_unknown, c_obs)
cand_dups = d2d < 5 * u.arcsec
mean, median, std = sigma_clipped_stats(image, sigma=3.0)
ap_area = CircularAperture((2, 2), r=aperture)
sig3_limiting_mag = -2.5 * np.log10(
3 * std * np.sqrt(ap_area.area())) + ZP_median
sig5_limiting_mag = -2.5 * np.log10(
5 * std * np.sqrt(ap_area.area())) + ZP_median
#######################
#estimating the error via random apertures
# print(std*np.sqrt(aperture_obj.area()))
ran_x = 2 * aperture + np.random.random(1000) * (image.shape[0] -
4 * aperture)
ran_y = 2 * aperture + np.random.random(1000) * (image.shape[0] -
4 * aperture)
apertures_error = CircularAperture((ran_x, ran_y), r=aperture)
phot_table = aperture_photometry(image, apertures_error)
phot_table[
'aperture_sum'].info.format = '%.8g' # for consistent table output
random_extractions = Table(phot_table).to_pandas()
#print(np.mean(np.abs(random_extractions["aperture_sum"])))
# print(np.median(np.abs(random_extractions["aperture_sum"])))
# sig5_limiting_mag_apertures = -2.5*np.log10(5*np.median(np.abs(random_extractions["aperture_sum"]))) + ZP_median
# print(sig5_limiting_mag_apertures)
m, _, std_apertures = sigma_clipped_stats(
random_extractions["aperture_sum"], sigma=3)
print(std_apertures)
sig5_limiting_mag_apertures = -2.5 * np.log10(
5 * std_apertures) + ZP_median
# print(sig5_limiting_mag_apertures)
# print(np.mean(random_extractions.loc[random_extractions["aperture_sum"]<0,"aperture_sum"]))
###########################
mag = 0
if (cand_dups.sum() > 0):
print(
"Position was given. Found {} sources in a 5 arcsec radius. Here is the magnitude:"
.format(cand_dups.sum()))
print(new_magnitudes[idx])
print("----")
aperture_obj = CircularAperture(
((observation["xcenter"].values)[idx],
(observation["ycenter"].values)[idx]),
r=aperture)
#noise = (observation["aperture_sum"].values)[idx] /(std*np.sqrt(aperture_obj.area()))
#https://en.wikipedia.org/wiki/Sum_of_normally_distributed_random_variables
#the std**2 is added, so we get a square root for the area!
noise = (
observation["aperture_sum"].values)[idx] / std_apertures
#print(noise_prob_wrong)
#print(noise_prob_wrong2)
print("We get a signal to noise of {} for the fixed aperture".
format(noise))
mag = new_magnitudes[idx]
mag_err = 2.5 * np.log10(1 + 1 / noise)
text = "found detection in {} band,\n {:.4g} +- {:.2g} {}mag, {:.3g} S/N \n 5 sig limiting mag is {:.4g}".format(
args.band, mag, mag_err, mag_sys, noise,
sig5_limiting_mag_apertures)
else:
print(
"Position was given. No object was found at that position."
)
pix_unknown = wcsprm.s2p([[ra, dec]], 1)
pix_unknown = pix_unknown['pixcrd']
aperture_obj = CircularAperture(pix_unknown, r=aperture)
phot_table = aperture_photometry(image, aperture_obj)
phot_table[
'aperture_sum'].info.format = '%.8g' # for consistent table output
#print(phot_table)
forced_phot = Table(phot_table).to_pandas()
forced_mag = -2.5 * np.log10(
forced_phot["aperture_sum"].values) + ZP_median
forced_mag = forced_mag[0]
#noise = (forced_phot["aperture_sum"].values)[0] /(std*np.sqrt(aperture_obj.area()))
noise = (forced_phot["aperture_sum"].values)[0] / std_apertures
print("Forced photometry gives a magnitude of {}".format(
forced_mag))
#local error
# pix_unknown = pix_unknown[0]
# mean_loc, median_loc, std_loc = sigma_clipped_stats(image[int(pix_unknown[0])-50:int(pix_unknown[0])+50,int(pix_unknown[1])-50:int(pix_unknown[1])+50], sigma=3.0)
# ap_area= CircularAperture((2,2), r=aperture)
# sig5_limiting_mag_local = -2.5*np.log10(5*std_loc*np.sqrt(ap_area.area())) + ZP_median
# print((std_loc*np.sqrt(aperture_obj.area())))
# print(sig5_limiting_mag_local)
print("We get a signal to noise of {} for the fixed aperture".
format(noise))
mag = forced_mag
mag_err = 2.5 * np.log10(1 + 1 / noise)
text = "forced photometry in {} band,\n {:.4g} +- {:.2g} {}mag, {:.3g} S/N \n 5 sig limiting mag is {:.4g} ".format(
args.band, mag, mag_err, mag_sys, noise,
sig5_limiting_mag_apertures)
else:
print(new_magnitudes)
mag = 0
obs_with_photometry = observation.copy()
obs_with_photometry["mag"] = new_magnitudes
#search for targeted object and print out its magnitude TODO
if (ra and dec):
plt.figure(figsize=(15, 8))
plt.subplot(1, 2, 1)
pix_unknown = wcsprm.s2p([[ra, dec]], 1)
pix_unknown = pix_unknown['pixcrd']
pix_unknown = pix_unknown[0]
plt.xlabel("pixel x direction")
plt.ylabel("pixel y direction")
plt.title(text, size=20)
#plt.imshow(image[int( pix_unknown[0])-30: int(pix_unknown[0])+30, int(pix_unknown[1])-30:int(pix_unknown[1])+30],cmap='Greys', origin='lower', norm=LogNorm())
#plt.plot(30, 30, "+", color="red", markersize=20)
plt.imshow(image, cmap='Greys', origin='lower', norm=LogNorm())
plt.xlim(int(pix_unknown[0]) - 20, int(pix_unknown[0]) + 20)
plt.ylim(int(pix_unknown[1]) - 20, int(pix_unknown[1]) + 20)
aperture_obj.plot(color='blue',
lw=1.5,
alpha=0.5,
label=str(args.aperture) + " arcsec aperture")
plt.plot(pix_unknown[0],
pix_unknown[1],
"+",
color="red",
markersize=20,
linewidth=10,
label="predicted target position")
plt.legend(bbox_to_anchor=(1.1, -0.1), ncol=2)
plt.subplot(1, 2, 2)
plt.title("Photometric calibration with {} sources from {}".format(
obs_matched.shape[0], catalog_name))
plt.plot(cat_matched[band_name].values,
ZP,
"o",
markersize=20,
label="catalog objects")
if (MAG_CALC):
plt.ylim(np.nanmin(ZP) - 1, np.nanmax(ZP) + 1)
else:
plt.ylim(10, 30)
if (mag):
plt.axvline(x=mag, linewidth=4, color="green", label="target")
plt.xlabel("catalog " + band_name + " magnitude")
plt.ylabel(
"Zeropoint (should be constant in linear part of the detector)"
)
plt.legend()
outputname = fits_image_filename.replace('.fits', '')
plt.savefig(outputname + "_ra{}dec{}.pdf".format(ra, dec))
plt.show()
else:
plt.figure()
plt.plot(cat_matched[band_name].values,
ZP,
"o",
markersize=20,
label="catalog objects")
plt.ylim(np.nanmin(ZP) - 1, np.nanmax(ZP) + 1)
if (mag):
plt.axvline(x=mag, linewidth=4, color="green", label="target")
plt.xlabel("catalog " + band_name + " magnitude")
plt.ylabel(
"Zeropoint (should be constant in linear part of the detector)"
)
plt.legend()
plt.show()
print("overall time taken")
print(datetime.now() - StartTime)
# if(args.show_images):
# plt.show()
print("-- finished --")