fix negative stride (torch); fix inferring nv in icwt

ssqueezepy/_cwt.py

@@ -1,7 +1,7 @@
 # -*- coding: utf-8 -*-
 import numpy as np
 from .utils import fft, ifft, ifftshift, FFT_GLOBAL
-from .utils import WARN, adm_cwt, adm_ssq, _process_fs_and_t
+from .utils import WARN, adm_cwt, adm_ssq, _process_fs_and_t, is_array_or_tensor
 from .utils import padsignal, process_scales, logscale_transition_idx
 from .utils import backend as S
 from .utils.backend import Q
@@ -404,7 +404,7 @@ def icwt(Wx, wavelet='gmw', scales='log-piecewise', nv=None, one_int=True,
     #### Prepare for inversion ###############################################
     *_, na, n = Wx.shape
     x_len = x_len or n
-    if not isinstance(scales, np.ndarray) and nv is None:
+    if not is_array_or_tensor(scales) and nv is None:
         nv = 32  # must match forward's; default to `cwt`'s
 
     wavelet = _process_gmw_wavelet(wavelet, l1_norm)

ssqueezepy/utils/backend.py

@@ -46,6 +46,11 @@ def is_tensor(*args, mode='all'):
     return cond(isinstance(x, torch.Tensor) for x in args)
 
 
+def is_array_or_tensor(*args, mode='all'):
+    cond = all if mode == 'all' else any
+    return cond(isinstance(x, (torch.Tensor, np.ndarray)) for x in args)
+
+
 def is_dtype(x, str_dtype):
     return (str_dtype in str(x.dtype) if isinstance(str_dtype, str) else
             any(sd in str(x.dtype) for sd in str_dtype))

ssqueezepy/utils/cwt_utils.py

@@ -588,7 +588,7 @@ def integrate_analytic(int_fn, nowarn=False):
     Integrates near zero separately in log space (useful for e.g. 1/x).
     """
     def _est_arr(mxlim, N):
-        t = np.linspace(mxlim, .1, N, endpoint=False)[::-1]
+        t = np.linspace(mxlim, .1, N, endpoint=False)[::-1].copy()
         arr = int_fn(t)
 
         max_idx = np.argmax(arr)