Periodogram¶

This tutorial demonstrates the Lomb–Scargle periodogram via light_curve.Periodogram.

Synthetic periodic light curve¶

In [ ]:

import light_curve as lc
import numpy as np

rng = np.random.default_rng(0)
PERIOD = 23.5  # days
t = np.sort(rng.uniform(0, 300, 150))
m = 15.0 + 0.4 * np.sin(2 * np.pi * t / PERIOD) + rng.normal(0, 0.04, 150)
err = np.full(150, 0.04)
print(f'True period: {PERIOD} days, n_obs: {len(t)}')

import light_curve as lc import numpy as np rng = np.random.default_rng(0) PERIOD = 23.5 # days t = np.sort(rng.uniform(0, 300, 150)) m = 15.0 + 0.4 * np.sin(2 * np.pi * t / PERIOD) + rng.normal(0, 0.04, 150) err = np.full(150, 0.04) print(f'True period: {PERIOD} days, n_obs: {len(t)}')

True period: 23.5 days, n_obs: 150

Basic usage¶

Periodogram(peaks=1) returns the best period and its normalised power:

In [ ]:

pg = lc.Periodogram(peaks=1, fast=True)
result = pg(t, m, err)
for name, val in zip(pg.names, result):
    print(f'  {name}: {val:.4f}')

pg = lc.Periodogram(peaks=1, fast=True) result = pg(t, m, err) for name, val in zip(pg.names, result): print(f' {name}: {val:.4f}')

Multiple peaks¶

peaks=5 returns the five strongest periods:

In [ ]:

pg5 = lc.Periodogram(peaks=5)
result5 = pg5(t, m, err)
print('Top 5 periods (days):')
for i in range(5):
    print(f'  peak {i+1}: period = {result5[i*2]:.3f} d, power = {result5[i*2+1]:.4f}')

pg5 = lc.Periodogram(peaks=5) result5 = pg5(t, m, err) print('Top 5 periods (days):') for i in range(5): print(f' peak {i+1}: period = {result5[i*2]:.3f} d, power = {result5[i*2+1]:.4f}')

Notes¶

• fast=True (default) uses an FFT-based algorithm; set fast=False for the exact O(N²) implementation.
• Periods are in the same units as the input time array.
• See API reference for all constructor options.