""" NUFFT HSA classes (deprecated) ======================================= """ from __future__ import absolute_import import numpy import warnings import scipy.sparse import numpy.fft import scipy.signal import scipy.linalg import scipy.special from functools import wraps as _wraps from ..src._helper import helper, helper1 class hypercube: def __init__(self, shape, steps, invsteps, nelements, batch, dtype): self.shape = shape self.steps = steps self.invsteps = invsteps self.nelements = nelements self.batch = batch self.dtype = dtype def push_cuda_context(hsa_method): """ Decorator to push up CUDA context to the top of the stack for current use Add @push_cuda_context before the methods of NUFFT_hsa() """ @_wraps(hsa_method) def wrapper(*args, **kwargs): try: args[0].thr._context.push() except: pass return hsa_method(*args, **kwargs) return wrapper class NUFFT_hsa: """ NUFFT_hsa class. Multi-coil or single-coil memory reduced NUFFT. """ def __init__(self, API=None, platform_number=None, device_number=None, verbosity=0): """ Constructor. :param API: The API for the heterogeneous system. API='cuda' or API='ocl' :param platform_number: The number of the platform found by the API. :param device_number: The number of the device found on the platform. :param verbosity: Defines the verbosity level, default value is 0 :type API: string :type platform_number: integer :type device_number: integer :type verbosity: integer :returns: 0 :rtype: int, float :Example: >>> from pynufft import NUFFT_hsa >>> NufftObj = NUFFT_hsa(API='cuda', platform_number=0, device_number=0, verbosity=0) """ warnings.warn('In the future NUFFT_hsa and NUFFT_cpu api will' ' be merged', FutureWarning) self.dtype = numpy.complex64 self.verbosity = verbosity import reikna.cluda as cluda if self.verbosity > 0: print('The choosen API by the user is ', API) self.cuda_flag, self.ocl_flag = helper.diagnose( verbosity=self.verbosity) if None is API: if self.cuda_flag is 1: API = 'cuda' elif self.ocl_flag is 1: API = 'ocl' else: warnings.warn('No parallelization will be made since no GPU ' 'device has been detected.', UserWarning) else: api = API if self.verbosity > 0: print('The used API will be ', API) if platform_number is None: platform_number = 0 if device_number is None: device_number = 0 from reikna import cluda import reikna.transformations from reikna.cluda import functions, dtypes try: # try to create api/platform/device using the given parameters if 'cuda' == API: api = cluda.cuda_api() elif 'ocl' == API: api = cluda.ocl_api() platform = api.get_platforms()[platform_number] device = platform.get_devices()[device_number] except: # if failed, find out what's going wrong? warnings.warn('No parallelization will be made since no GPU ' 'device has been detected.', UserWarning) # return 1 # Create context from device self.thr = api.Thread(device) # pyopencl.create_some_context() self.device = device # : device name if self.verbosity > 0: print('Using opencl or cuda = ', self.thr.api) # """ # Wavefront: as warp in cuda. Can control the width in a workgroup # Wavefront is required in spmv_vector as it improves data coalescence. # see cCSR_spmv and zSparseMatVec # """ self.wavefront = api.DeviceParameters(device).warp_size if self.verbosity > 0: print('Wavefront of OpenCL (as wrap of CUDA) = ', self.wavefront) from ..src import re_subroutine # import create_kernel_sets kernel_sets = re_subroutine.create_kernel_sets(API) prg = self.thr.compile(kernel_sets, render_kwds=dict(LL=str(self.wavefront)), fast_math=False) self.prg = prg def set_wavefront(self, wf): self.wavefront = int(wt)#api.DeviceParameters(device).warp_size if self.verbosity > 0: print('Wavefront of OpenCL (as wrap of CUDA) = ', self.wavefront) from ..src import re_subroutine # import create_kernel_sets kernel_sets = re_subroutine.create_kernel_sets(API) prg = self.thr.compile(kernel_sets, render_kwds=dict(LL=str(self.wavefront)), fast_math=False) self.prg = prg def plan(self, om, Nd, Kd, Jd, ft_axes=None, batch=None, radix=None): """ Design the multi-coil or single-coil memory reduced interpolator. :param om: The M off-grid locations in the frequency domain. Normalized between [-pi, pi] :param Nd: The matrix size of equispaced image. Example: Nd=(256, 256) for a 2D image; Nd = (128, 128, 128) for a 3D image :param Kd: The matrix size of the oversampled frequency grid. Example: Kd=(512,512) for 2D image; Kd = (256,256,256) for a 3D image :param Jd: The interpolator size. Example: Jd=(6,6) for 2D image; Jd = (6,6,6) for a 3D image :param ft_axes: The dimensions to be transformed by FFT. Example: ft_axes = (0, 1) for 2D, ft_axes = (0, 1, 2) for 3D; ft_axes = None for all dimensions. :param batch: Batch NUFFT. If provided, the shape is Nd + (batch, ). The last axis is the number of parallel coils. batch = None for single coil. :param radix: ????. If provided, the shape is Nd + (batch, ). The last axis is the number of parallel coils. batch = None for single coil. :type om: numpy.float array, matrix size = (M, ndims) :type Nd: tuple, ndims integer elements. :type Kd: tuple, ndims integer elements. :type Jd: tuple, ndims integer elements. :type ft_axes: tuple, selected axes to be transformed. :type batch: int or None :returns: 0 :rtype: int, float :Example: >>> from pynufft import NUFFT_hsa >>> NufftObj = NUFFT_hsa() >>> NufftObj.plan(om, Nd, Kd, Jd) """ self.ndims = len(Nd) # dimension if ft_axes is None: ft_axes = range(0, self.ndims) self.ft_axes = ft_axes self.st = helper.plan(om, Nd, Kd, Jd, ft_axes=ft_axes, format='pELL', radix=radix) if batch is None: self.parallel_flag = 0 else: self.parallel_flag = 1 if batch is None: self.batch = numpy.uint32(1) else: self.batch = numpy.uint32(batch) self.Nd = self.st['Nd'] # backup self.Kd = self.st['Kd'] # self.sn = numpy.asarray(self.st['sn'].astype(self.dtype), # order='C')# backup if self.batch == 1 and (self.parallel_flag == 0): self.multi_Nd = self.Nd self.multi_Kd = self.Kd self.multi_M = (self.st['M'], ) # Broadcasting the sense and scaling factor (Roll-off) # self.sense2 = self.sense*numpy.reshape(self.sn, self.Nd + (1, )) else: # self.batch is 0: self.multi_Nd = self.Nd + (self.batch, ) self.multi_Kd = self.Kd + (self.batch, ) self.multi_M = (self.st['M'], ) + (self.batch, ) self.invbatch = 1.0 / self.batch self.Kdprod = numpy.uint32(numpy.prod(self.st['Kd'])) self.Jdprod = numpy.uint32(numpy.prod(self.st['Jd'])) self.Ndprod = numpy.uint32(numpy.prod(self.st['Nd'])) self.Nd_elements, self.invNd_elements = helper.strides_divide_itemsize( self.st['Nd']) # only return the Kd_elements self.Kd_elements = helper.strides_divide_itemsize(self.st['Kd'])[0] self.NdCPUorder, self.KdCPUorder, self.nelem = helper.preindex_copy( self.st['Nd'], self.st['Kd']) self.offload() return 0 @push_cuda_context def offload(self): # API, platform_number=0, device_number=0): """ self.offload(): Off-load NUFFT to the opencl or cuda device(s) :param API: define the device type, which can be 'cuda' or 'ocl' :param platform_number: define which platform to be used. The default platform_number = 0. :param device_number: define which device to be used. The default device_number = 0. :type API: string :type platform_number: int :type device_number: int :return: self: instance """ self.pELL = {} # dictionary self.pELL['nRow'] = numpy.uint32(self.st['pELL'].nRow) self.pELL['prodJd'] = numpy.uint32(self.st['pELL'].prodJd) self.pELL['sumJd'] = numpy.uint32(self.st['pELL'].sumJd) self.pELL['dim'] = numpy.uint32(self.st['pELL'].dim) self.pELL['Jd'] = self.thr.to_device( self.st['pELL'].Jd.astype(numpy.uint32)) self.pELL['meshindex'] = self.thr.to_device( self.st['pELL'].meshindex.astype(numpy.uint32)) self.pELL['kindx'] = self.thr.to_device( self.st['pELL'].kindx.astype(numpy.uint32)) self.pELL['udata'] = self.thr.to_device( self.st['pELL'].udata.astype(self.dtype)) self.volume = {} self.volume['Nd_elements'] = self.thr.to_device( numpy.asarray(self.Nd_elements, dtype=numpy.uint32)) self.volume['Kd_elements'] = self.thr.to_device( numpy.asarray(self.Kd_elements, dtype=numpy.uint32)) self.volume['invNd_elements'] = self.thr.to_device( self.invNd_elements.astype(numpy.float32)) self.volume['Nd'] = self.thr.to_device(numpy.asarray( self.st['Nd'], dtype=numpy.uint32)) self.volume['NdGPUorder'] = self.thr.to_device(self.NdCPUorder) self.volume['KdGPUorder'] = self.thr.to_device(self.KdCPUorder) self.volume['gpu_coil_profile'] = self.thr.array( self.multi_Nd, dtype=self.dtype).fill(1.0) Nd = self.st['Nd'] # tensor_sn = numpy.empty((numpy.sum(Nd), ), dtype=numpy.float32) # # shift = 0 # for dimid in range(0, len(Nd)): # # tensor_sn[shift :shift + Nd[dimid]] = \ # self.st['tensor_sn'][dimid][:, 0].real # shift = shift + Nd[dimid] # self.volume['tensor_sn'] = self.thr.to_device( # self.st['tensor_sn'].astype(numpy.float32)) self.tSN = {} self.tSN['Td_elements'] = self.thr.to_device( numpy.asarray(self.st['tSN'].Td_elements, dtype=numpy.uint32)) self.tSN['invTd_elements'] = self.thr.to_device( self.st['tSN'].invTd_elements.astype(numpy.float32)) self.tSN['Td'] = self.thr.to_device( numpy.asarray(self.st['tSN'].Td, dtype=numpy.uint32)) self.tSN['Tdims'] = self.st['tSN'].Tdims self.tSN['tensor_sn'] = self.thr.to_device( self.st['tSN'].tensor_sn.astype(numpy.float32)) self.Ndprod = numpy.int32(numpy.prod(self.st['Nd'])) self.Kdprod = numpy.int32(numpy.prod(self.st['Kd'])) self.M = numpy.int32(self.st['M']) import reikna.fft if self.batch > 1: # batch mode self.fft = reikna.fft.FFT( numpy.empty(self.st['Kd']+(self.batch, ), dtype=self.dtype), self.ft_axes).compile(self.thr, fast_math=False) else: # elf.Reps == 1 Batch mode is wrong for self.fft = reikna.fft.FFT( numpy.empty(self.st['Kd'], dtype=self.dtype), self.ft_axes).compile(self.thr, fast_math=False) self.zero_scalar = self.dtype(0.0+0.0j) del self.st['pELL'] if self.verbosity > 0: print('End of offload') @push_cuda_context def reset_sense(self): self.volume['gpu_coil_profile'].fill(1.0) @push_cuda_context def set_sense(self, coil_profile): if coil_profile.shape != self.multi_Nd: print('The shape of coil_profile is ', coil_profile.shape) print('But it should be', self.Nd + (self.batch, )) raise ValueError else: self.volume['gpu_coil_profile'] = self.thr.to_device( coil_profile.astype(self.dtype)) if self.verbosity > 0: print('Successfully loading coil sensitivities!') # if coil_profile.shape == self.Nd + (self.batch, ): @push_cuda_context def to_device(self, image, shape=None): g_image = self.thr.array(image.shape, dtype=self.dtype) self.thr.to_device(image.astype(self.dtype), dest=g_image) return g_image @push_cuda_context def s2x(self, s): x = self.thr.array(self.multi_Nd, dtype=self.dtype) # print("Now populate the array to multi-coil") self.prg.cPopulate( self.batch, self.Ndprod, s, x, local_size=None, global_size=int(self.batch * self.Ndprod)) # x2 = x * self.volume['gpu_coil_profile'] # try: # x2 = x * self.volume['gpu_coil_profile'] # except: # x2 = x self.prg.cMultiplyVecInplace( numpy.uint32(1), self.volume['gpu_coil_profile'], x, local_size=None, global_size=int(self.batch*self.Ndprod)) # self.prg.cDistribute( # self.batch, # self.Ndprod, # self.volume['gpu_coil_profile'], # s, # x, # local_size=None, # global_size=int(self.batch*self.Ndprod)) return x @push_cuda_context def x2xx(self, x): # xx = self.thr.array(xx.shape, dtype = self.dtype) # self.thr.copy_array(z, dest=xx, ) # size = int(xx.nbytes/xx.dtype.itemsize)) # Hack of error in cuda backends; 8 is the byte of numpy.complex64 # size = int(xx.nbytes/8) xx = self.thr.array(x.shape, dtype=self.dtype) self.thr.copy_array(x, dest=xx, ) # size = int(xx.nbytes/xx.dtype.itemsize)) # Hack of error in cuda backends; 8 is the byte of numpy.complex64: # size = int(xx.nbytes/8) # self.prg.cMultiplyRealInplace( # self.batch, # self.volume['SnGPUArray'], # xx, # local_size=None, # global_size=int(self.Ndprod*self.batch)) # self.prg.cTensorMultiply(numpy.uint32(self.batch), # numpy.uint32(self.ndims), # self.volume['Nd'], # self.volume['Nd_elements'], # self.volume['invNd_elements'], # self.volume['tensor_sn'], # xx, # numpy.uint32(0), # local_size=None, # global_size=int(self.batch*self.Ndprod)) self.prg.cTensorMultiply(numpy.uint32(self.batch), numpy.uint32(self.tSN['Tdims']), self.tSN['Td'], self.tSN['Td_elements'], self.tSN['invTd_elements'], self.tSN['tensor_sn'], xx, numpy.uint32(0), local_size=None, global_size=int(self.batch*self.Ndprod)) # self.thr.synchronize() return xx @push_cuda_context def xx2k(self, xx): """ Private: oversampled FFT on the heterogeneous device First, zeroing the self.k_Kd array Second, copy self.x_Nd array to self.k_Kd array by cSelect Third, inplace FFT """ k = self.thr.array(self.multi_Kd, dtype=self.dtype) # k = self.thr.array(self.multi_Kd, dtype=self.dtype).fill(0.0 + 0.0j) k.fill(0) # self.prg.cMultiplyScalar(self.zero_scalar, # k, # local_size=None, # global_size=int(self.Kdprod)) # # self.prg.cSelect(self.NdGPUorder, # self.KdGPUorder, # xx, # k, # local_size=None, # global_size=int(self.Ndprod)) # self.prg.cSelect2(self.batch, # self.volume['NdGPUorder'], # self.volume['KdGPUorder'], # xx, # k, # local_size=None, # global_size=int(self.Ndprod*self.batch)) self.prg.cTensorCopy( self.batch, numpy.uint32(self.ndims), self.volume['Nd_elements'], self.volume['Kd_elements'], self.volume['invNd_elements'], xx, k, numpy.int32(1), # Directions: Nd -> Kd, 1; Kd -> Nd, -1 local_size=None, global_size=int(self.Ndprod)) self.fft(k, k, inverse=False) # self.thr.synchronize() return k @push_cuda_context def k2y(self, k): """ Private: interpolation by the Sparse Matrix-Vector Multiplication """ # if self.parallel_flag is 1: # y =self.thr.array((self.st['M'], self.batch), # dtype=self.dtype).fill(0) # else: # y =self.thr.array( (self.st['M'], ), dtype=self.dtype).fill(0) y = self.thr.array(self.multi_M, dtype=self.dtype).fill(0) self.prg.pELL_spmv_mCoil( self.batch, self.pELL['nRow'], self.pELL['prodJd'], self.pELL['sumJd'], self.pELL['dim'], self.pELL['Jd'], # self.pELL_currsumJd, self.pELL['meshindex'], self.pELL['kindx'], self.pELL['udata'], k, y, local_size=int(self.wavefront), global_size=int(self.pELL['nRow'] * self.batch * self.wavefront) ) # self.thr.synchronize() return y @push_cuda_context def y2k(self, y): """ Private: gridding by the Sparse Matrix-Vector Multiplication However, serial atomic add is far too slow and inaccurate. """ # kx = self.thr.array(self.multi_Kd, dtype=numpy.float32).fill(0.0) # ky = self.thr.array(self.multi_Kd, dtype=numpy.float32).fill(0.0) k = self.thr.array(self.multi_Kd, dtype=numpy.complex64).fill(0.0) res = self.thr.array(self.multi_Kd, dtype=numpy.complex64).fill(0.0) # array which saves the residue of two sum self.prg.pELL_spmvh_mCoil( self.batch, self.pELL['nRow'], self.pELL['prodJd'], self.pELL['sumJd'], self.pELL['dim'], self.pELL['Jd'], self.pELL['meshindex'], self.pELL['kindx'], self.pELL['udata'], # kx, ky, k, res, y, local_size=None, global_size=int(self.pELL['nRow']*self.batch )#* # int(self.pELL['prodJd']) * int(self.batch)) ) return k + res @push_cuda_context def k2xx(self, k): """ Private: the inverse FFT and image cropping (which is the reverse of _xx2k() method) """ self.fft(k, k, inverse=True) # self.thr.synchronize() # self.x_Nd._zero_fill() # self.prg.cMultiplyScalar(self.zero_scalar, # xx, # local_size=None, # global_size=int(self.Ndprod)) # if self.parallel_flag is 1: # xx = self.thr.array(self.st['Nd']+(self.batch, ), # dtype = self.dtype) # else: # xx = self.thr.array(self.st['Nd'], dtype = self.dtype) xx = self.thr.array(self.multi_Nd, dtype=self.dtype) xx.fill(0) # self.prg.cSelect(self.queue, # (self.Ndprod,), # None, # self.volume['KdGPUorder'].data, # self.NdGPUorder.data, # self.k_Kd2.data, # self.x_Nd.data) # self.prg.cSelect2(self.batch, # self.volume['KdGPUorder'], # self.volume['NdGPUorder'], # k, # xx, # local_size=None, # global_size=int(self.Ndprod*self.batch)) self.prg.cTensorCopy( self.batch, numpy.uint32(self.ndims), self.volume['Nd_elements'], self.volume['Kd_elements'], self.volume['invNd_elements'], k, xx, numpy.int32(-1), local_size=None, global_size=int(self.Ndprod)) return xx @push_cuda_context def xx2x(self, xx): x = self.thr.array(xx.shape, dtype=self.dtype) self.thr.copy_array(xx, dest=x, ) # size = int(xx.nbytes/xx.dtype.itemsize)) # Hack of error in cuda backends; 8 is the byte of numpy.complex64 # size = int(xx.nbytes/8) # self.prg.cMultiplyRealInplace(self.batch, # self.volume['SnGPUArray'], # z, # local_size=None, # global_size=int(self.batch*self.Ndprod)) # self.prg.cTensorMultiply(numpy.uint32(self.batch), # numpy.uint32(self.ndims), # self.volume['Nd'], # self.volume['Nd_elements'], # self.volume['invNd_elements'], # self.volume['tensor_sn'], # x, # numpy.uint32(0), # local_size=None, # global_size=int(self.batch*self.Ndprod)) self.prg.cTensorMultiply(numpy.uint32(self.batch), numpy.uint32(self.tSN['Tdims']), self.tSN['Td'], self.tSN['Td_elements'], self.tSN['invTd_elements'], self.tSN['tensor_sn'], x, numpy.uint32(0), local_size=None, global_size=int(self.batch * self.Ndprod)) # self.thr.synchronize() return x @push_cuda_context def x2s(self, x): s = self.thr.array(self.st['Nd'], dtype=self.dtype) # try: self.prg.cMultiplyConjVecInplace( numpy.uint32(1), self.volume['gpu_coil_profile'], x, local_size=None, global_size=int(self.batch*self.Ndprod)) # x2 = x * self.volume['gpu_coil_profile'].conj() # except: # x2 = x self.prg.cAggregate( self.batch, self.Ndprod, x, s, local_size=int(self.wavefront), global_size=int(self.batch*self.Ndprod*self.wavefront)) # self.prg.cMerge(self.batch, # self.Ndprod, # self.volume['gpu_coil_profile'], # x, # s, # local_size=int(self.wavefront), # global_size = int(self.batch*self.Ndprod* # self.wavefront)) return s @push_cuda_context def selfadjoint_one2many2one(self, gx): """ selfadjoint_one2many2one NUFFT on the heterogeneous device :param gx: The input gpu array, with size=Nd :type gx: reikna gpu array with dtype =numpy.complex64 :return: gx: The output gpu array, with size=Nd :rtype: reikna gpu array with dtype =numpy.complex64 """ gy = self.forward_one2many(gx) gx2 = self.adjoint_many2one(gy) del gy return gx2 def selfadjoint(self, gx): """ selfadjoint NUFFT on the heterogeneous device :param gx: The input gpu array, with size=Nd :type gx: reikna gpu array with dtype =numpy.complex64 :return: gx: The output gpu array, with size=Nd :rtype: reikna gpu array with dtype =numpy.complex64 """ gy = self.forward(gx) gx2 = self.adjoint(gy) del gy return gx2 @push_cuda_context def forward(self, gx): """ Forward NUFFT on the heterogeneous device :param gx: The input gpu array, with size = Nd :type gx: reikna gpu array with dtype = numpy.complex64 :return: gy: The output gpu array, with size = (M,) :rtype: reikna gpu array with dtype = numpy.complex64 """ try: xx = self.x2xx(gx) except: # gx is not a gpu array try: warnings.warn('The input array may not be a GPUarray ' 'Automatically moving the input array to gpu, ' 'which is throttled by PCIe.', UserWarning) px = self.to_device(gx, ) # pz = self.thr.to_device(numpy.asarray(gz.astype(self.dtype), # order = 'C' )) xx = self.x2xx(px) except: if gx.shape != self.Nd + (self.batch, ): warnings.warn('Shape of the input is ' + str(gx.shape) + ' while it should be ' + str(self.Nd+(self.batch, )), UserWarning) raise k = self.xx2k(xx) del xx gy = self.k2y(k) del k return gy @push_cuda_context def forward_one2many(self, s): try: x = self.s2x(s) except: # gx is not a gpu array try: warnings.warn('In s2x(): The input array may not be ' 'a GPUarray. Automatically moving the input' ' array to gpu, which is throttled by PCIe.', UserWarning) ps = self.to_device(s, ) # px = self.thr.to_device(numpy.asarray(x.astype(self.dtype), # order = 'C' )) x = self.s2x(ps) except: if s.shape != self.Nd: warnings.warn('Shape of the input is ' + str(x.shape) + ' while it should be ' + str(self.Nd), UserWarning) raise y = self.forward(x) return y @push_cuda_context def adjoint_many2one(self, y): try: x = self.adjoint(y) except: # gx is not a gpu array try: if self.verbosity > 0: print('y.shape = ', y.shape) warnings.warn('In adjoint_many2one(): The input array may not ' 'be a GPUarray. Automatically moving the input' ' array to gpu, which is throttled by PCIe.', UserWarning) py = self.to_device(y, ) # py = self.thr.to_device(numpy.asarray(y.astype(self.dtype), # order = 'C' )) x = self.adjoint(py) except: print('Failed at self.adjoint_many2one! Please check the gy' ' shape, type and stride.') raise # z = self.adjoint(y) s = self.x2s(x) return s @push_cuda_context def adjoint(self, gy): """ Adjoint NUFFT on the heterogeneous device :param gy: The input gpu array, with size=(M,) :type: reikna gpu array with dtype =numpy.complex64 :return: gx: The output gpu array, with size=Nd :rtype: reikna gpu array with dtype =numpy.complex64 """ try: k = self.y2k(gy) except: # gx is not a gpu array try: warnings.warn('In adjoint(): The input array may not ' 'be a GPUarray. Automatically moving the input' ' array to gpu, which is throttled by PCIe.', UserWarning) py = self.to_device(gy, ) # py = self.thr.to_device(numpy.asarray(gy.astype(self.dtype), # order = 'C' )) k = self.y2k(py) except: print('Failed at self.adjont! Please check the gy shape, ' 'type, stride.') raise # k = self.y2k(gy) xx = self.k2xx(k) del k gx = self.xx2x(xx) del xx return gx @push_cuda_context def release(self): del self.volume del self.prg del self.pELL self.thr.release() del self.thr @push_cuda_context def solve(self, gy, solver=None, *args, **kwargs): """ The solver of NUFFT_hsa :param gy: data, reikna array, (M,) size :param solver: could be 'cg', 'L1TVOLS', 'L1TVLAD' :param maxiter: the number of iterations :type gy: reikna array, dtype = numpy.complex64 :type solver: string :type maxiter: int :return: reikna array with size Nd """ from ..linalg.solve_hsa import solve try: return solve(self, gy, solver, *args, **kwargs) except: try: warnings.warn('In solve(): The input array may not ' 'be a GPUarray. Automatically moving the input' ' array to gpu, which is throttled by PCIe.', UserWarning) py = self.to_device(gy, ) return solve(self, py, solver, *args, **kwargs) except: if numpy.ndarray == type(gy): print("Input gy must be a reikna array with dtype =" " numpy.complex64") raise # TypeError else: print("wrong") raise # TypeError