def test_bin_uncertainty(self):
data = [[[[[[100, 100], [100, 100]], 'G']], [141.42, 141.42],
'simple'],
[[[[[100, 100], [100, 200]], 'G']], [141.42, 223.61],
'simple 2'],
[[[[[100, 100], [100, 100]], 'L'],
[[[100, 100], [100, 100]], 'L']], [100, 100], 'simple 3'],
[[[[[200, 200], [100, 100]], 'L'],
[[[200, 100], [100, 100]], 'L']], [158.11, 132.29],
'simple 4'],
[[[[[200, 200], [100, 100]], 'L'],
[[[200, 100], [100, 100]], 'A']], [158.11, 132.29],
'weigh makes no difference if equal'],
[[[[[100, 100], [100, 100]], 'L']
for i in range(4)] + [[[[100, 100], [100, 100]], 'G']],
[66.67, 66.67], 'one correction'],
[[[[[100, 100], [100, 100]], 'L'],
[[[100, 100], [100, 100]], 'G']], [105.41, 105.41],
'two corrections'], [[], [0, 0], 'empty']]
for i in data:
g_actual = wh.g_actual_bins(i[0])
for j in [1, 2]:
res = wh.bin_uncertainty(self.g_exp, g_actual, i[0], j, 0)
self.assertAlmostEqual(res[0], i[1][0] / j, msg=i[2], places=2)
self.assertAlmostEqual(res[1], i[1][1] / j, msg=i[2], places=2)
def test_g_actual_bins(self):
data = [[[[[1, 2], 'G']], {
'G': 1,
'A': 0,
'L': 0
}], [[[[1, 2], 'G'], [[1, 2], 'A']], {
'G': 0.5,
'A': 0.5,
'L': 0
}],
[[[[1, 2], 'G'], [[1, 2], 'A'], [[1, 2], 'G']], {
'G': 2 / 3,
'A': 1 / 3,
'L': 0
}],
[[[[1, 2], 'G'], [[1, 2], 'A'], [[1, 2], 'G'], [[1, 2], 'L']],
{
'G': 0.5,
'A': 0.25,
'L': 0.25
}], [[], {
'A': 0,
'G': 0,
'L': 0
}]]
for i in data:
self.assertEqual(wh.g_actual_bins(i[0]), i[1])
def sf_m_at_z(bin_maxes, index, t_step):
locs = ['A', 'L', 'G'] #ALG or a subset
if locs != ['A', 'L', 'G']:
print('WARNING, only using {}'.format(locs))
sfh = data.sfh()
masses, masses_unc = sfh.abs_mass()
g_loc = sfh.get_loc()
z0, t0, z1, t1 = h.z_and_t_from_index_t_step(index, t_step, sfh.z_times)
gal_mass = [[galaxy[index], galaxy[index + t_step]] for galaxy in masses]
gal_mass_unc = [[galaxy[index], galaxy[index + t_step]]
for galaxy in masses_unc]
print(z0)
# Bin each gal, throw error if not binned or smaller than first bin
orig_bins, orig_bins_unc = [[[] for i in bin_maxes] for j in range(2)]
for j, gal in enumerate(gal_mass): # for each galaxy
if g_loc[j] not in locs:
continue
for i in range(len(bin_maxes)
): # scan through bins. Bin and break when appropriate
if math.log10(gal[0]) < bin_maxes[i]:
orig_bins[i].append([gal, g_loc[j]])
orig_bins_unc[i].append([gal_mass_unc[j], g_loc[j]])
break
else:
raise Exception("Check max bin sizing")
if orig_bins[0]:
raise Exception("Check min bin sizing")
orig_bins, orig_bins_unc = orig_bins[1:], orig_bins_unc[1:]
for i in orig_bins:
print(i)
print(len(i))
print('')
sys.exit()
bins, bins_unc, bins_center, bins_center_unc = [[] for i in range(4)]
for i in range(len(orig_bins)):
g_actual = wh.g_actual_bins(orig_bins[i])
bins.append(
wh.average_growth_rate(sfh.g_exp, g_actual, orig_bins[i], t0, t1))
bins_unc.append(
wh.bin_uncertainty(sfh.g_exp, g_actual, orig_bins_unc[i], t0, t1))
bins_center.append(
wh.average_bin_start(sfh.g_exp, g_actual, orig_bins[i]))
bins_center_unc.append(
wh.uncertainty_bin_start(sfh.g_exp, g_actual, orig_bins_unc[i],
bins_center[-1]))
return (bins, bins_unc, bins_center, bins_center_unc)
def test_average_bin_start(self):
data = [[[[[100, 888], 'G']], 2, 'simple'],
[[[[80, 777], 'G'], [[120, 444], 'G']], 2,
'expected freqs make no diff when only one g type'],
[[[[80, 999], 'A'], [[120, 555], 'L']], 2,
'expected freqs make no diff when gs of same prob'],
[[[[100, 333], 'A'], [[150, 111], 'G']], 2.07,
'expected freqs correct bin when acutal != exp'],
[[], 0, 'empty']]
for i in data:
g_actual = wh.g_actual_bins(i[0])
self.assertAlmostEqual(wh.average_bin_start(
self.g_exp, g_actual, i[0]),
i[1],
msg=i[2],
places=2)
def test_uncertainty_bin_start(self):
data = [[[[[[100, 200], [777, 888]], 'G']], 3, [0.0414, 0.0969],
'simple'],
[[[[[100, 200], [777, 888]], 'G']], 4, [0.0043, 0.0088],
'simple'],
[[[[[100, 100], [100, 100]], 'L']
for i in range(4)] + [[[[100, 100], [100, 100]], 'G']], 3,
[0.0200, 0.0210], 'one correction'],
[[[[[100, 100], [100, 100]], 'L'],
[[[100, 100], [100, 100]], 'G']], 2, [0.2419, 0.5941],
'two corrections'], [[], 3, [0, 0], 'empty']]
for i in data:
g_actual = wh.g_actual_bins(i[0])
res = wh.uncertainty_bin_start(self.g_exp, g_actual, i[0], i[1])
self.assertAlmostEqual(res[0], i[2][0], msg=i[3], places=4)
self.assertAlmostEqual(res[1], i[2][1], msg=i[3], places=4)
def test_average_growth_rate(self):
data = [[[[[100, 200], 'G']], 100, 'simple'],
[[[[100, 200], 'G'], [[150, 200], 'G']], 75,
'expected freqs make no diff when only one g type'],
[[[[100, 200], 'A'], [[150, 200], 'L']], 75,
'expected freqs make no diff when gs of same prob'],
[[[[100, 200], 'A'], [[150, 200], 'G']], 250 / 3,
'expected freqs correct bin when acutal != exp'],
[[], 0, 'empty']]
for i in data:
g_actual = wh.g_actual_bins(i[0])
self.assertAlmostEqual(wh.average_growth_rate(
self.g_exp, g_actual, i[0], 1, 0),
i[1],
msg=i[2])
self.assertAlmostEqual(wh.average_growth_rate(
self.g_exp, g_actual, i[0], 2, 0),
i[1] / 2,
msg=i[2])