lanczos (pyleoclim.utils.filter.lanczos)

pyleoclim.utils.filter.lanczos(ys, fc, fs=1, pad='reflect', reflect_type='odd', params=(1, 0, 0), padFrac=0.1)[source]

Applies a Lanczos (lowpass) filter with frequency fc, with optional padding

Parameters

  • ys (numpy array) – Timeseries

  • fc (float) – cutoff frequency.

  • fs (float) – sampling frequency

  • pad (string) – Indicates if padding is needed. - ‘reflect’: Reflects the timeseries - ‘ARIMA’: Uses an ARIMA model for the padding - None: No padding.

  • params (tuple) – model parameters for ARIMA model (if pad = ‘ARIMA’). May require fiddling.

  • padFrac (float) – fraction of the series to be padded

Returns

yf – filtered array

Return type

array

References

Filter design from http://scitools.org.uk/iris/docs/v1.2/examples/graphics/SOI_filtering.html

See also

pyleoclim.utils.filter.ts_pad

Pad a timeseries based on timeseries model predictions