make_freq_vector (pyleoclim.utils.wavelet.make_freq_vector)

pyleoclim.utils.wavelet.make_freq_vector(ts, method='log', **kwargs)[source]

Make frequency vector

This function selects among five methods to obtain the frequency vector.

Parameters

  • ts (array) – Time axis of the time series

  • method (string) – The method to use. Options are ‘log’ (default), ‘nfft’, ‘lomb_scargle’, ‘welch’, and ‘scale’

  • kwargs (dict, optional) –

    For Lomb_Scargle, additional parameters may be passed: - nf (int): number of frequency points - ofac (float): Oversampling rate that influences the resolution of the frequency axis,

    when equals to 1, it means no oversamling (should be >= 1). The default value 4 is usaually a good value.

    • hifac (float): fhi/fnyq (should be >= 1), where fhi is the highest frequency that

      can be analyzed by the Lomb-Scargle algorithm and fnyq is the Nyquist frequency.

Returns

freq – the frequency vector

Return type

array

See also

pyleoclim.utils.wavelet.freq_vector_lomb_scargle

Return the frequency vector based on the REDFIT recommendation.

pyleoclim.utils.wavelet.freq_vector_welch

Return the frequency vector based on the Welch’s method.

pyleoclim.utils.wavelet.freq_vector_nfft

Return the frequency vector based on NFFT

pyleoclim.utils.wavelet.freq_vector_scale

Return the frequency vector based on scales

pyleoclim.utils.wavelet.freq_vector_log

Return the frequency vector based on logspace