pyhctsa.operations.wavelet.cwt¶
-
pyhctsa.operations.wavelet.cwt(y, w_name=
'db3', max_scale=32)¶ Continuous wavelet transform of a time series.
Computes the continuous wavelet transform (CWT) of a time series using the specified wavelet and maximum scale, and returns a set of statistics on the coefficients, entropy, and results of coefficients summed across scales.
- Parameters:¶
- y : array-like¶
The input time series.
- w_name : str, optional¶
The wavelet name, e.g., ‘db3’ (Daubechies wavelet), ‘sym2’ (Symlet), or the continuous Morlet wavelet ‘morl’ (the second wavelet registered in hctsa). For either kind, the wavelet function is integrated on a grid of 2**10 points over its support, and the integral, stretched to each scale, is convolved with the series and differenced (the continuous transform with that wavelet at integer scales). Default is
'db3'.- max_scale : int, optional¶
The maximum scale of wavelet analysis. Default is 32.
- power : The scaled power is relative to the mean
- with : SC = S / mean(S)
- 1 : energy shares of the strongest
- of : independent of the number
- pover99 : coefficients).
- pover98
- pover95
- the : 10 and 20% of coefficients (the sum of
- 1
- 2
- 5
- the
- ``max(1 : largest
- SC : floor((100 - p)/100*numEntries))`` values of
- by : divided
- maximum : numEntries). SC_h is the entropy of SC relative to its
- Return type:¶
dict
:param : :param
-sum(SC_a*log(SC_a)) - log(numEntries)withSC_a = SC/sum(SC).:- Returns:¶
Dictionary of statistics on the CWT coefficients, including
gam1andgam2(the shape and scale of a gamma distribution fitted to SC by maximum likelihood; NaN if any scaled power is exactly zero, which the gamma distribution cannot describe) anddd_SC_h(the entropy of the maximum of SC in each of 10 time boxes at each scale).- Return type:¶
dict
- Parameters:¶