deepextractor.generation.generate_timeseries¶

Generate synthetic time-domain training data.

Usage:

deepextractor-generate --output-dir data/ --num-train 250000 --bilby-noise

Module Contents¶

deepextractor.generation.generate_timeseries.SAMPLE_RATE = 4096[source]¶

deepextractor.generation.generate_timeseries.T = 2.0[source]¶

deepextractor.generation.generate_timeseries.T_INJ = 1.0[source]¶

deepextractor.generation.generate_timeseries.LENGTH = 0[source]¶

deepextractor.generation.generate_timeseries.MINIMUM_FREQUENCY = 20.0[source]¶

deepextractor.generation.generate_timeseries.SNR_SCALING_FACTOR_BILBY = 31.970149253731343[source]¶

deepextractor.generation.generate_timeseries.SIGNAL_TYPES = ['chirp', 'sine', 'sine_gaussian', 'gaussian_pulse', 'ringdown'][source]¶

deepextractor.generation.generate_timeseries.SIGNAL_FUNCTION_MAP[source]¶

deepextractor.generation.generate_timeseries.generate_gaussian_noise(mean, std_dev, num_samples, sample_shape, bilby_noise=False, sample_rate=SAMPLE_RATE, duration=T, minimum_frequency=MINIMUM_FREQUENCY, detector='L1')[source]¶: Generate Gaussian noise samples (pycbc or bilby).

deepextractor.generation.generate_timeseries.generate_synthetic_data(gaussian_noise_samples, bilby_noise=False, phase='train', t_min=0.125, t_max=2.0, snr_min=SNR_MIN, snr_max=SNR_MAX)[source]¶: Generate synthetic noisy glitch and background data arrays.

deepextractor.generation.generate_timeseries.main()[source]¶