Fix incompatibility with scipy 1.14.0 and numpy 2.0

sigpy/alg.py

@@ -3,6 +3,7 @@
 and implements commonly used methods, such as gradient methods,
 Newton's method, and the augmented Lagrangian method.
 """
+
 import numpy as np
 
 import sigpy as sp
@@ -88,7 +89,7 @@ class PowerMethod(Alg):
     def __init__(self, A, x, norm_func=None, max_iter=30):
         self.A = A
         self.x = x
-        self.max_eig = np.infty
+        self.max_eig = np.inf
         self.norm_func = norm_func
         super().__init__(max_iter)
 
@@ -180,7 +181,7 @@ def __init__(
                 self.z = self.x.copy()
                 self.t = 1
 
-        self.resid = np.infty
+        self.resid = np.inf
         super().__init__(max_iter)
 
     def _update(self):
@@ -371,7 +372,7 @@ def __init__(
             with self.u_device:
                 self.sigma_min = xp.amin(xp.abs(sigma)).item()
 
-        self.resid = np.infty
+        self.resid = np.inf
 
         super().__init__(max_iter)
 
@@ -490,7 +491,7 @@ def _update(self):
             xp = device.xp
             with device:
                 util.axpy(self.u, self.mu, self.g(self.x))
-                backend.copyto(self.u, xp.clip(self.u, 0, np.infty))
+                backend.copyto(self.u, xp.clip(self.u, 0, np.inf))
 
         if self.h is not None:
             util.axpy(self.v, self.mu, self.h(self.x))
@@ -802,10 +803,10 @@ def __init__(
         self.gradf = gradf
         self.inv_hessf = inv_hessf
         self.x = x
-        self.lamda = np.infty
+        self.lamda = np.inf
         self.beta = beta
         self.f = f
-        self.residual = np.infty
+        self.residual = np.inf
         self.tol = tol
 
         super().__init__(max_iter)
@@ -864,7 +865,7 @@ def __init__(self, A, y, x0, max_iter=500, tol=0, max_tol=0, lamb=0):
         self.tol = tol
         self.max_tol = max_tol
         self.lamb = lamb
-        self.residual = np.infty
+        self.residual = np.inf
 
     def _update(self):
         device = backend.get_device(self.y)

sigpy/mri/rf/slr.py

@@ -1,6 +1,6 @@
 # -*- coding: utf-8 -*-
 """MRI RF excitation pulse design functions,
-    including SLR and small tip spatial design
+including SLR and small tip spatial design
 """
 
 import numpy as np
@@ -135,7 +135,7 @@ def dzls(n=64, tb=4, d1=0.01, d2=0.01):
     m = [1, 1, 0, 0]
     w = [1, d1 / d2]
 
-    h = signal.firls(n + 1, f, m, w)
+    h = signal.firls(n + 1, f, m, weight=w)
     # shift the filter half a sample to make it symmetric, like in MATLAB
     c = np.exp(
         1j
@@ -160,7 +160,7 @@ def dzmp(n=64, tb=4, d1=0.01, d2=0.01):
     m = [1, 0]
     w = [1, 2 * d1 / (0.5 * d2 * d2)]
 
-    hl = signal.remez(n2, f, m, w)
+    hl = signal.remez(n2, f, m, weight=w)
 
     h = fmp(hl)
 
@@ -188,7 +188,7 @@ def dzlp(n=64, tb=4, d1=0.01, d2=0.01):
     m = [1, 0]
     w = [1, d1 / d2]
 
-    h = signal.remez(n, f, m, w)
+    h = signal.remez(n, f, m, weight=w)
 
     return h
 
@@ -253,8 +253,8 @@ def dz_gslider_b(
             m_sub = [1, 1, 0, 0, 0, 0]
             w = [1, 1, d1 / d2]
 
-            b_notch = signal.firls(n + 1, f, m_notch, w)  # the notched filter
-            b_sub = signal.firls(n + 1, f, m_sub, w)  # the subslice filter
+            b_notch = signal.firls(n + 1, f, m_notch, weight=w)  # the notched filter
+            b_sub = signal.firls(n + 1, f, m_sub, weight=w)  # the subslice filter
             # add them with the subslice phase
             b = np.add(b_notch, np.multiply(np.exp(1j * phi), b_sub))
             # shift the filter half a sample to make it symmetric,
@@ -344,15 +344,15 @@ def dz_gslider_b(
             )
         )
 
-        b_notch = signal.firls(n + 1, f, m_notch, w)  # the notched filter
+        b_notch = signal.firls(n + 1, f, m_notch, weight=w)  # the notched filter
         b_notch = sp.ifft(
             np.multiply(sp.fft(b_notch, center=False), c), center=False
         )
         b_notch = np.real(b_notch[:n])
         # hilbert transform to suppress negative passband
         b_notch = signal.hilbert(b_notch)
 
-        b_sub = signal.firls(n + 1, f, m_sub, w)  # the sub-band filter
+        b_sub = signal.firls(n + 1, f, m_sub, weight=w)  # the sub-band filter
         b_sub = sp.ifft(
             np.multiply(sp.fft(b_sub, center=False), c), center=False
         )
@@ -460,8 +460,8 @@ def dz_hadamard_b(n=128, g=5, gind=1, tb=4, d1=0.01, d2=0.01, shift=32):
                 [np.arange(0, n / 2 + 1, 1), np.arange(-n / 2, 0, 1)]
             )
         )
-        bp = signal.firls(n + 1, f, mp, w)  # the positive filter
-        bn = signal.firls(n + 1, f, mn, w)  # the negative filter
+        bp = signal.firls(n + 1, f, mp, weight=w)  # the positive filter
+        bn = signal.firls(n + 1, f, mn, weight=w)  # the negative filter
 
         # combine the filters and demodulate
         b = sp.ifft(
@@ -793,7 +793,7 @@ def dz_recursive_rf(
         win_len = (win_fact - 1) * n
         npad = n * z_pad_fact - win_fact * n
         # blackman window?
-        window = signal.blackman(int((win_fact - 1) * n))
+        window = signal.windows.blackman(int((win_fact - 1) * n))
         # split in half; stick N ones in the middle
         window = np.concatenate(
             (

sigpy/mri/rf/trajgrad.py

@@ -906,7 +906,7 @@ def sdotmax(
     ds_p = np.linalg.norm(cp1_highres, axis=1)
 
     # s vs p to enable conversion
-    s_of_p = integrate.cumtrapz(ds_p, p_highres, initial=0)
+    s_of_p = integrate.cumulative_trapezoid(ds_p, p_highres, initial=0)
     curve_length = s_of_p[-1]
 
     # decide ds and compute st for the first point
@@ -981,7 +981,7 @@ def sdotmax(
     st_of_s = np.minimum(sta, stb)
 
     # compute time
-    t_of_s = integrate.cumtrapz(1.0 / st_of_s, initial=0) * ds
+    t_of_s = integrate.cumulative_trapezoid(1.0 / st_of_s, initial=0) * ds
 
     t = np.arange(0, t_of_s[-1] + np.finfo(float).eps, dt)
 
@@ -992,7 +992,7 @@ def sdotmax(
     g = np.diff(c, axis=0, append=np.zeros((1, 3))) / gamma / dt
     g[-1, :] = g[-2, :] + g[-2, :] - g[-3, :]
 
-    k = integrate.cumtrapz(g, t, initial=0, axis=0) * gamma
+    k = integrate.cumulative_trapezoid(g, t, initial=0, axis=0) * gamma
 
     s = np.diff(g, axis=0) / dt
 

tests/test_version.py

@@ -8,4 +8,4 @@
 
 class TestVersion(unittest.TestCase):
     def test_version(self):
-        assert version.__version__ == "0.1.25"
+        assert version.__version__ == "0.1.26"