I did seasonal decomposition on my time series using statsmodels library in Python. How can I get the periods/frequencies of the different subsignals composing the observed signal ?
Is there a function giving the Fourier decomposition of the seasonal signal ?
The seasonal package (PyPI, or github) includes a seasonal.periodogram command that will plot the significant periodicities in your data, with or without first detrending (it uses this to estimate the major period of your data).
seasonal.periodogram uses Welch's method (no relation to me), which handles noisy input signals more robustly than a pure Fourier Transform. It uses the SciPy implementation in scipy.signal.welch
