def __call__(self, fname, mapfiles):
"""
Consolidate a list of mapfiles, and delete input mapfiles
if clear_intermediate_files is True.
Parameters
----------
fname : `str`
Output filename
mapfiles : `list`
Input list of files
"""
print("Consolidating %d maps into %s" % (len(mapfiles), fname))
healsparse.cat_healsparse_files(mapfiles, fname)
if self.clear_intermediate_files:
for f in mapfiles:
os.unlink(f)
def test_cat_maps(self):
self.test_dir = tempfile.mkdtemp(dir='./', prefix='TestHealSparse-')
# Make 3 maps of the same type. These will have different nside_coverage
# to exercise the up-scale and down-scaling. The output nside_coverage
# will be in the middle
nside_coverage1 = 32
nside_coverage2 = 64
nside_coverage3 = 128
nside_sparse = 1024
nside_coverage_out = 64
for t in ['array', 'recarray', 'widemask']:
if t == 'array':
dtype = np.float64
primary = None
wide_mask_maxbits = None
sentinel = hpg.UNSEEN
data = np.array([100.0], dtype=dtype)
elif t == 'recarray':
dtype = [('a', 'f4'),
('b', 'i4')]
primary = 'a'
wide_mask_maxbits = None
sentinel = hpg.UNSEEN
data = np.zeros(1, dtype=dtype)
data['a'] = 100.0
data['b'] = 100
else:
dtype = healsparse.WIDE_MASK
primary = None
wide_mask_maxbits = 24
sentinel = 0
map1 = healsparse.HealSparseMap.make_empty(nside_coverage1, nside_sparse,
dtype, primary=primary,
sentinel=sentinel,
wide_mask_maxbits=wide_mask_maxbits)
if t == 'widemask':
map1.set_bits_pix(np.arange(10000, 15000), [1, 12])
else:
map1[10000: 15000] = data
map2 = healsparse.HealSparseMap.make_empty(nside_coverage2, nside_sparse,
dtype, primary=primary,
sentinel=sentinel,
wide_mask_maxbits=wide_mask_maxbits)
if t == 'widemask':
map2.set_bits_pix(np.arange(15001, 20000), [1, 12])
else:
map2[15001: 20000] = data
map3 = healsparse.HealSparseMap.make_empty(nside_coverage3, nside_sparse,
dtype, primary=primary,
sentinel=sentinel,
wide_mask_maxbits=wide_mask_maxbits)
if t == 'widemask':
map3.set_bits_pix(np.arange(20001, 25000), [1, 12])
else:
map3[20001: 25000] = data
# Make a combined map for comparison
# This should have the same nside_coverage as map2
combined_map = healsparse.HealSparseMap.make_empty_like(map1,
nside_coverage=nside_coverage_out)
combined_map[map1.valid_pixels] = map1[map1.valid_pixels]
combined_map[map2.valid_pixels] = map2[map2.valid_pixels]
combined_map[map3.valid_pixels] = map3[map3.valid_pixels]
filename1 = os.path.join(self.test_dir, 'test_%s_1.hs' % (t))
filename2 = os.path.join(self.test_dir, 'test_%s_2.hs' % (t))
filename3 = os.path.join(self.test_dir, 'test_%s_3.hs' % (t))
map1.write(filename1)
map2.write(filename2)
map3.write(filename3)
file_list = [filename1, filename2, filename3]
for in_mem in [False, True]:
outfile = os.path.join(self.test_dir, 'test_%s_combined_%d.hs' %
(t, int(in_mem)))
if not healsparse.fits_shim.use_fitsio and not in_mem:
# We cannot use out-of-memory option with astropy.io.fits
self.assertRaises(RuntimeError, cat_healsparse_files,
file_list, outfile, in_memory=in_mem,
nside_coverage_out=nside_coverage_out)
else:
cat_healsparse_files(file_list, outfile, in_memory=in_mem,
nside_coverage_out=nside_coverage_out)
map_test = healsparse.HealSparseMap.read(outfile)
testing.assert_array_equal(map_test.valid_pixels, combined_map.valid_pixels)
if t == 'recarray':
for col in map_test.dtype.names:
testing.assert_array_almost_equal(map_test[:][col],
combined_map[:][col])
else:
testing.assert_array_almost_equal(map_test[:], combined_map[:])