def test_percent_close_to_median():
"""Test feature which finds the percentage of points near the median value."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['percent_close_to_median'])
amplitude = (max(values) - min(values)) / 2.
within_buffer = np.abs(values - np.median(values)) < 0.2 * amplitude
npt.assert_allclose(f['percent_close_to_median'], np.mean(within_buffer))
def test_percent_close_to_median():
"""Test feature which finds the percentage of points near the median value."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors,
['percent_close_to_median'])
amplitude = (max(values) - min(values)) / 2.
within_buffer = np.abs(values - np.median(values)) < 0.2*amplitude
npt.assert_allclose(f['percent_close_to_median'], np.mean(within_buffer))
def test_delta_t_hist():
"""Test histogram of all time lags."""
times, values, errors = irregular_random(500)
delta_ts = [pair[1] - pair[0] for pair in itertools.combinations(times, 2)]
nbins = 50
bins = np.linspace(0, max(times) - min(times), nbins + 1)
npt.assert_allclose(cf.delta_t_hist(times, nbins),
np.histogram(delta_ts, bins=bins)[0], atol=2)
def test_scatter_res_raw():
"""Test feature that measures scatter of Lomb-Scargle residuals."""
times, values, errors = irregular_random()
lomb_model = lomb_scargle.lomb_scargle_model(times, values, errors)
residuals = values - lomb_model['freq_fits'][0]['model']
resid_mad = np.median(np.abs(residuals - np.median(residuals)))
value_mad = np.median(np.abs(values - np.median(values)))
f = generate_features(times, values, errors, ['scatter_res_raw'])
npt.assert_allclose(f['scatter_res_raw'], resid_mad / value_mad, atol=3e-2)
def test_scatter_res_raw():
"""Test feature that measures scatter of Lomb-Scargle residuals."""
times, values, errors = irregular_random()
lomb_model = lomb_scargle.lomb_scargle_model(times, values, errors)
residuals = values - lomb_model['freq_fits'][0]['model']
resid_mad = np.median(np.abs(residuals - np.median(residuals)))
value_mad = np.median(np.abs(values - np.median(values)))
f = generate_features(times, values, errors, ['scatter_res_raw'])
npt.assert_allclose(f['scatter_res_raw'], resid_mad / value_mad, atol=3e-2)
def test_amplitude():
"""Test features related to amplitude/magnitude percentiles."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['amplitude'])
npt.assert_allclose(f['amplitude'], (max(values) - min(values)) / 2.)
f = generate_features(times, values, errors, ['percent_amplitude'])
max_scaled = 10**(-0.4 * max(values))
min_scaled = 10**(-0.4 * min(values))
med_scaled = 10**(-0.4 * np.median(values))
peak_from_median = max(abs((max_scaled - med_scaled) / med_scaled),
abs((min_scaled - med_scaled)) / med_scaled)
npt.assert_allclose(f['percent_amplitude'], peak_from_median, rtol=5e-4)
f = generate_features(times, values, errors,
['percent_difference_flux_percentile'])
band_offset = 13.72
w_m2 = 10**(-0.4 * (values + band_offset) - 3
) # 1 erg/s/cm^2 = 10^-3 w/m^2
npt.assert_allclose(
f['percent_difference_flux_percentile'],
np.diff(np.percentile(w_m2, [5, 95])) / np.median(w_m2))
f = generate_features(times, values, errors,
['flux_percentile_ratio_mid20'])
npt.assert_allclose(
f['flux_percentile_ratio_mid20'],
np.diff(np.percentile(w_m2, [40, 60])) /
np.diff(np.percentile(w_m2, [5, 95])))
f = generate_features(times, values, errors,
['flux_percentile_ratio_mid35'])
npt.assert_allclose(
f['flux_percentile_ratio_mid35'],
np.diff(np.percentile(w_m2, [32.5, 67.5])) /
np.diff(np.percentile(w_m2, [5, 95])))
f = generate_features(times, values, errors,
['flux_percentile_ratio_mid50'])
npt.assert_allclose(
f['flux_percentile_ratio_mid50'],
np.diff(np.percentile(w_m2, [25, 75])) /
np.diff(np.percentile(w_m2, [5, 95])))
f = generate_features(times, values, errors,
['flux_percentile_ratio_mid65'])
npt.assert_allclose(
f['flux_percentile_ratio_mid65'],
np.diff(np.percentile(w_m2, [17.5, 82.5])) /
np.diff(np.percentile(w_m2, [5, 95])))
f = generate_features(times, values, errors,
['flux_percentile_ratio_mid80'])
npt.assert_allclose(
f['flux_percentile_ratio_mid80'],
np.diff(np.percentile(w_m2, [10, 90])) /
np.diff(np.percentile(w_m2, [5, 95])))
def test_delta_t_hist():
"""Test histogram of all time lags."""
times, values, errors = irregular_random(500)
delta_ts = [pair[1] - pair[0] for pair in itertools.combinations(times, 2)]
nbins = 50
bins = np.linspace(0, max(times) - min(times), nbins + 1)
npt.assert_allclose(cf.delta_t_hist(times, nbins),
np.histogram(delta_ts, bins=bins)[0],
atol=2)
def test_normalize_hist():
"""Test normalization of histogram."""
times, values, errors = irregular_random(500)
delta_ts = [pair[1] - pair[0] for pair in itertools.combinations(times, 2)]
nbins = 50
bins = np.linspace(0, max(times) - min(times), nbins + 1)
nhist = cf.normalize_hist(cf.delta_t_hist(times, nbins),
max(times) - min(times))
npt.assert_allclose(nhist, np.histogram(delta_ts, bins=bins,
density=True)[0], atol=0.01)
def test_qso_features():
"""Test features which measure fit of QSO model.
Reference values are hard-coded values from previous implementation; not sure
of examples with a closed-form solution.
"""
times, values, errors = irregular_random()
f = generate_features(times, values, errors,
['qso_log_chi2_qsonu', 'qso_log_chi2nuNULL_chi2nu'])
npt.assert_allclose(f['qso_log_chi2_qsonu'], 6.9844064754)
npt.assert_allclose(f['qso_log_chi2nuNULL_chi2nu'], -0.456526327522)
def test_qso_features():
"""Test features which measure fit of QSO model.
Reference values are hard-coded values from previous implementation; not sure
of examples with a closed-form solution.
"""
times, values, errors = irregular_random()
f = generate_features(times, values, errors,
['qso_log_chi2_qsonu', 'qso_log_chi2nuNULL_chi2nu'])
npt.assert_allclose(f['qso_log_chi2_qsonu'], 6.9844064754)
npt.assert_allclose(f['qso_log_chi2nuNULL_chi2nu'], -0.456526327522)
def test_normalize_hist():
"""Test normalization of histogram."""
times, values, errors = irregular_random(500)
delta_ts = [pair[1] - pair[0] for pair in itertools.combinations(times, 2)]
nbins = 50
bins = np.linspace(0, max(times) - min(times), nbins + 1)
nhist = cf.normalize_hist(cf.delta_t_hist(times, nbins),
max(times) - min(times))
npt.assert_allclose(nhist,
np.histogram(delta_ts, bins=bins, density=True)[0],
atol=0.01)
def test_weighted_average():
"""Test weighted average and distance from weighted average features."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['weighted_average'])
weighted_avg = np.average(values, weights=1. / (errors**2))
weighted_var = np.average((values - weighted_avg)**2,
weights=1. / (errors**2))
npt.assert_allclose(f['weighted_average'], weighted_avg)
dists_from_weighted_avg = values - weighted_avg
stds_from_weighted_avg = (dists_from_weighted_avg / np.sqrt(weighted_var))
f = generate_features(times, values, errors, ['percent_beyond_1_std'])
npt.assert_equal(f['percent_beyond_1_std'],
np.mean(stds_from_weighted_avg > 1.))
def test_weighted_average():
"""Test weighted average and distance from weighted average features."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['weighted_average'])
weighted_avg = np.average(values, weights=1. / (errors**2))
weighted_var = np.average((values - weighted_avg)**2,
weights=1. / (errors**2))
npt.assert_allclose(f['weighted_average'], weighted_avg)
dists_from_weighted_avg = values - weighted_avg
stds_from_weighted_avg = (dists_from_weighted_avg /
np.sqrt(weighted_var))
f = generate_features(times, values, errors,
['percent_beyond_1_std'])
npt.assert_equal(f['percent_beyond_1_std'], np.mean(stds_from_weighted_avg > 1.))
def test_stetson():
"""Test Stetson variability features."""
times, values, errors = irregular_random(size=201)
f = generate_features(times, values, errors, ['stetson_j'])
# Stetson mean approximately equal to standard mean for large inputs
dists = np.sqrt(float(len(values)) / (len(values) - 1.)) * (values - np.mean(values)) / 0.1
npt.assert_allclose(f['stetson_j'],
np.mean(np.sign(dists**2-1)*np.sqrt(np.abs(dists**2-1))),
rtol=1e-2)
# Stetson_j should be somewhat close to (scaled) variance for normal data
npt.assert_allclose(f['stetson_j']*0.1, np.var(values), rtol=2e-1)
# Hard-coded original value
npt.assert_allclose(f['stetson_j'], 7.591347175195703)
f = generate_features(times, values, errors, ['stetson_k'])
npt.assert_allclose(f['stetson_k'], 1./0.798 * np.mean(np.abs(dists)) / np.sqrt(np.mean(dists**2)), rtol=5e-4)
# Hard-coded original value
npt.assert_allclose(f['stetson_k'], 1.0087218792719013)
def test_amplitude():
"""Test features related to amplitude/magnitude percentiles."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['amplitude'])
npt.assert_allclose(f['amplitude'], (max(values) - min(values)) /2.)
f = generate_features(times, values, errors, ['percent_amplitude'])
max_scaled = 10**(-0.4 * max(values))
min_scaled = 10**(-0.4 * min(values))
med_scaled = 10**(-0.4 * np.median(values))
peak_from_median = max(abs((max_scaled - med_scaled) / med_scaled),
abs((min_scaled - med_scaled)) / med_scaled)
npt.assert_allclose(f['percent_amplitude'], peak_from_median, rtol=5e-4)
f = generate_features(times, values, errors, ['percent_difference_flux_percentile'])
band_offset = 13.72
w_m2 = 10**(-0.4*(values+band_offset)-3) # 1 erg/s/cm^2 = 10^-3 w/m^2
npt.assert_allclose(f['percent_difference_flux_percentile'], np.diff(
np.percentile(w_m2, [5, 95])) / np.median(w_m2))
f = generate_features(times, values, errors, ['flux_percentile_ratio_mid20'])
npt.assert_allclose(f['flux_percentile_ratio_mid20'],
np.diff(np.percentile(w_m2, [40, 60])) /
np.diff(np.percentile(w_m2, [5, 95])))
f = generate_features(times, values, errors, ['flux_percentile_ratio_mid35'])
npt.assert_allclose(f['flux_percentile_ratio_mid35'],
np.diff(np.percentile(w_m2, [32.5, 67.5])) /
np.diff(np.percentile(w_m2, [5, 95])))
f = generate_features(times, values, errors, ['flux_percentile_ratio_mid50'])
npt.assert_allclose(f['flux_percentile_ratio_mid50'],
np.diff(np.percentile(w_m2, [25, 75])) /
np.diff(np.percentile(w_m2, [5, 95])))
f = generate_features(times, values, errors, ['flux_percentile_ratio_mid65'])
npt.assert_allclose(f['flux_percentile_ratio_mid65'],
np.diff(np.percentile(w_m2, [17.5, 82.5])) /
np.diff(np.percentile(w_m2, [5, 95])))
f = generate_features(times, values, errors, ['flux_percentile_ratio_mid80'])
npt.assert_allclose(f['flux_percentile_ratio_mid80'],
np.diff(np.percentile(w_m2, [10, 90])) /
np.diff(np.percentile(w_m2, [5, 95])))
def test_stetson():
"""Test Stetson variability features."""
times, values, errors = irregular_random(size=201)
f = generate_features(times, values, errors, ['stetson_j'])
# Stetson mean approximately equal to standard mean for large inputs
dists = np.sqrt(float(len(values)) /
(len(values) - 1.)) * (values - np.mean(values)) / 0.1
npt.assert_allclose(
f['stetson_j'],
np.mean(np.sign(dists**2 - 1) * np.sqrt(np.abs(dists**2 - 1))),
rtol=1e-2)
# Stetson_j should be somewhat close to (scaled) variance for normal data
npt.assert_allclose(f['stetson_j'] * 0.1, np.var(values), rtol=2e-1)
# Hard-coded original value
npt.assert_allclose(f['stetson_j'], 7.591347175195703)
f = generate_features(times, values, errors, ['stetson_k'])
npt.assert_allclose(f['stetson_k'],
1. / 0.798 * np.mean(np.abs(dists)) /
np.sqrt(np.mean(dists**2)),
rtol=5e-4)
# Hard-coded original value
npt.assert_allclose(f['stetson_k'], 1.0087218792719013)
def test_std():
"""Test standard deviation feature."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['std'])
npt.assert_allclose(f['std'], np.std(values))
def test_skew():
"""Test statistical skew feature."""
from scipy import stats
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['skew'])
npt.assert_allclose(f['skew'], stats.skew(values))
def test_max_slope():
"""Test maximum slope feature, which finds the INDEX of the largest slope."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['max_slope'])
slopes = np.diff(values) / np.diff(times)
npt.assert_allclose(f['max_slope'], np.max(np.abs(slopes)))
def test_min():
"""Test minimum value feature."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['minimum'])
npt.assert_equal(f['minimum'], min(values))
def test_median():
"""Test median value feature."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['median'])
npt.assert_allclose(f['median'], np.median(values))
def test_median_absolute_deviation():
"""Test median absolute deviation (from the median) feature."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['median_absolute_deviation'])
npt.assert_allclose(f['median_absolute_deviation'],
np.median(np.abs(values - np.median(values))))
def test_median_absolute_deviation():
"""Test median absolute deviation (from the median) feature."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['median_absolute_deviation'])
npt.assert_allclose(f['median_absolute_deviation'], np.median(np.abs(values -
np.median(values))))
def test_max_slope():
"""Test maximum slope feature, which finds the INDEX of the largest slope."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['max_slope'])
slopes = np.diff(values) / np.diff(times)
npt.assert_allclose(f['max_slope'], np.max(np.abs(slopes)))
def test_median():
"""Test median value feature."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['median'])
npt.assert_allclose(f['median'], np.median(values))
def test_min():
"""Test minimum value feature."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['minimum'])
npt.assert_equal(f['minimum'], min(values))
def test_skew():
"""Test statistical skew feature."""
from scipy import stats
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['skew'])
npt.assert_allclose(f['skew'], stats.skew(values))
def test_std():
"""Test standard deviation feature."""
times, values, errors = irregular_random()
f = generate_features(times, values, errors, ['std'])
npt.assert_allclose(f['std'], np.std(values))
