docs: update py getting started to use cupy

Simpler.

docs/python_gpu.rst

@@ -7,15 +7,12 @@ Quick-start examples
 As mentioned in the :ref:`Python GPU installation instructions <install-python-gpu>`, the easiest way to install the Python interface for cuFINUFFT is to run ``pip install cufinufft``.
 
 Assuming cuFINUFFT has been installed, we will now consider how to calculate a 1D type 1 transform.
-To manage the GPU and transfer to and from host and device, we will use the ``pycuda`` library.
+To manage the GPU and transfer to and from host and device, we will use the ``cupy`` library.
 Consequently, we start with a few import statements.
 
 .. code-block:: python
 
-    import numpy as np
-
-    import pycuda.autoinit
-    from pycuda.gpuarray import to_gpu
+    import cupy as cp
 
     import cufinufft
 
@@ -30,26 +27,26 @@ We then proceed to setting up a few parameters.
     N = 200000
 
     # generate positions for the nonuniform points and the coefficients
-    x = 2 * np.pi * np.random.uniform(size=M)
-    c = (np.random.standard_normal(size=M)
-         + 1J * np.random.standard_normal(size=M))
+    x_gpu = 2 * cp.pi * cp.random.uniform(size=M)
+    c_gpu = (cp.random.standard_normal(size=M)
+             + 1J * cp.random.standard_normal(size=M))
 
 Now that the data is prepared, we need to set up a cuFINUFFT plan that can be executed on that data.
 
 .. code-block:: python
 
     # create plan
-    plan = cufinufft.Plan(1, (N,), dtype=np.float64)
+    plan = cufinufft.Plan(1, (N,), dtype="complex128")
 
     # set the nonuniform points
-    plan.setpts(to_gpu(x))
+    plan.setpts(x_gpu)
 
 With everything set up, we are now ready to execute the plan.
 
 .. code-block:: python
 
     # execute the plan
-    f_gpu = plan.execute(to_gpu(c))
+    f_gpu = plan.execute(c_gpu)
 
     # move results off the GPU
     f = f_gpu.get()