a possible bug in parameter 'freq'

[thelightonmyway]

Hi,

I use wavelet_coherence recently, and I encounter an error when I run:
freq=freq_vector_scale(time,dj=1/12,s0=2,j1=84,mother='MORLET',param=6)
coh=pv_series.wavelet_coherence(SWS_series, freq=freq)

and it shows:
ValueError: negative dimensions are not allowed

Something similar will happen here when I run:
coh=pv_series.wavelet_coherence(SWS_series, method='wwz',freq='scale')

I suppose it may be a bug that freq set as in reverse order, and it could be seen also in #626.

For reference, my parameter freq is set as:
time=np.arange(708) freq=freq_vector_scale(time,dj=1/12,s0=2,j1=84,mother='MORLET',param=6)

and it is:
array([0.48400665, 0.45684145, 0.43120091, 0.40699946, 0.38415634, 0.3625953 , 0.34224439, 0.32303569, 0.30490509, 0.28779208, 0.27163955, 0.25639359, 0.24200333, 0.22842072, 0.21560045, 0.20349973, 0.19207817, 0.18129765, 0.17112219, 0.16151784, 0.15245254, 0.14389604, 0.13581978, 0.1281968 , 0.12100166, 0.11421036, 0.10780023, 0.10174987, 0.09603908, 0.09064882, 0.0855611 , 0.08075892, 0.07622627, 0.07194802, 0.06790989, 0.0640984 , 0.06050083, 0.05710518, 0.05390011, 0.05087493, 0.04801954, 0.04532441, 0.04278055, 0.04037946, 0.03811314, 0.03597401, 0.03395494, 0.0320492 , 0.03025042, 0.02855259, 0.02695006, 0.02543747, 0.02400977, 0.02266221, 0.02139027, 0.02018973, 0.01905657, 0.017987 , 0.01697747, 0.0160246 , 0.01512521, 0.0142763 , 0.01347503, 0.01271873, 0.01200489, 0.0113311 , 0.01069514, 0.01009487, 0.00952828, 0.0089935 , 0.00848874, 0.0080123 , 0.0075626 , 0.00713815, 0.00673751, 0.00635937, 0.00600244, 0.00566555, 0.00534757, 0.00504743, 0.00476414, 0.00449675, 0.00424437, 0.00400615, 0.0037813 ])

The code won't work until the freq is rearrange in order:
coh=pv_series.wavelet_coherence(SWS_series,freq=freq[::-1])

Looking forward to your reply!