[eafc9fa] | 1 | """ |
---|
| 2 | Python driver for python kernels |
---|
| 3 | |
---|
| 4 | Calls the kernel with a vector of $q$ values for a single parameter set. |
---|
| 5 | Polydispersity is supported by looping over different parameter sets and |
---|
| 6 | summing the results. The interface to :class:`PyModel` matches those for |
---|
| 7 | :class:`kernelcl.GpuModel` and :class:`kerneldll.DllModel`. |
---|
| 8 | """ |
---|
[b3f6bc3] | 9 | import numpy as np |
---|
[c85db69] | 10 | from numpy import pi, cos |
---|
[b3f6bc3] | 11 | |
---|
[c85db69] | 12 | from .generate import F64 |
---|
[b3f6bc3] | 13 | |
---|
[f734e7d] | 14 | class PyModel(object): |
---|
[eafc9fa] | 15 | """ |
---|
| 16 | Wrapper for pure python models. |
---|
| 17 | """ |
---|
[17bbadd] | 18 | def __init__(self, model_info): |
---|
| 19 | self.info = model_info |
---|
[3c56da87] | 20 | |
---|
[eafc9fa] | 21 | def __call__(self, q_vectors): |
---|
| 22 | q_input = PyInput(q_vectors, dtype=F64) |
---|
| 23 | kernel = self.info['Iqxy'] if q_input.is_2d else self.info['Iq'] |
---|
[3c56da87] | 24 | return PyKernel(kernel, self.info, q_input) |
---|
| 25 | |
---|
[f734e7d] | 26 | def release(self): |
---|
[eafc9fa] | 27 | """ |
---|
| 28 | Free resources associated with the model. |
---|
| 29 | """ |
---|
[f734e7d] | 30 | pass |
---|
| 31 | |
---|
[b3f6bc3] | 32 | class PyInput(object): |
---|
| 33 | """ |
---|
| 34 | Make q data available to the gpu. |
---|
| 35 | |
---|
| 36 | *q_vectors* is a list of q vectors, which will be *[q]* for 1-D data, |
---|
| 37 | and *[qx, qy]* for 2-D data. Internally, the vectors will be reallocated |
---|
| 38 | to get the best performance on OpenCL, which may involve shifting and |
---|
| 39 | stretching the array to better match the memory architecture. Additional |
---|
| 40 | points will be evaluated with *q=1e-3*. |
---|
| 41 | |
---|
| 42 | *dtype* is the data type for the q vectors. The data type should be |
---|
| 43 | set to match that of the kernel, which is an attribute of |
---|
| 44 | :class:`GpuProgram`. Note that not all kernels support double |
---|
| 45 | precision, so even if the program was created for double precision, |
---|
| 46 | the *GpuProgram.dtype* may be single precision. |
---|
| 47 | |
---|
| 48 | Call :meth:`release` when complete. Even if not called directly, the |
---|
| 49 | buffer will be released when the data object is freed. |
---|
| 50 | """ |
---|
| 51 | def __init__(self, q_vectors, dtype): |
---|
| 52 | self.nq = q_vectors[0].size |
---|
| 53 | self.dtype = dtype |
---|
[eafc9fa] | 54 | self.is_2d = (len(q_vectors) == 2) |
---|
[c85db69] | 55 | self.q_vectors = [np.ascontiguousarray(q, self.dtype) for q in q_vectors] |
---|
[750ffa5] | 56 | self.q_pointers = [q.ctypes.data for q in self.q_vectors] |
---|
[b3f6bc3] | 57 | |
---|
| 58 | def release(self): |
---|
[eafc9fa] | 59 | """ |
---|
| 60 | Free resources associated with the model inputs. |
---|
| 61 | """ |
---|
[b3f6bc3] | 62 | self.q_vectors = [] |
---|
| 63 | |
---|
| 64 | class PyKernel(object): |
---|
| 65 | """ |
---|
| 66 | Callable SAS kernel. |
---|
| 67 | |
---|
| 68 | *kernel* is the DllKernel object to call. |
---|
| 69 | |
---|
[17bbadd] | 70 | *model_info* is the module information |
---|
[b3f6bc3] | 71 | |
---|
[3c56da87] | 72 | *q_input* is the DllInput q vectors at which the kernel should be |
---|
[b3f6bc3] | 73 | evaluated. |
---|
| 74 | |
---|
| 75 | The resulting call method takes the *pars*, a list of values for |
---|
| 76 | the fixed parameters to the kernel, and *pd_pars*, a list of (value,weight) |
---|
| 77 | vectors for the polydisperse parameters. *cutoff* determines the |
---|
| 78 | integration limits: any points with combined weight less than *cutoff* |
---|
| 79 | will not be calculated. |
---|
| 80 | |
---|
| 81 | Call :meth:`release` when done with the kernel instance. |
---|
| 82 | """ |
---|
[17bbadd] | 83 | def __init__(self, kernel, model_info, q_input): |
---|
| 84 | self.info = model_info |
---|
[3c56da87] | 85 | self.q_input = q_input |
---|
| 86 | self.res = np.empty(q_input.nq, q_input.dtype) |
---|
[eafc9fa] | 87 | dim = '2d' if q_input.is_2d else '1d' |
---|
[b3f6bc3] | 88 | # Loop over q unless user promises that the kernel is vectorized by |
---|
| 89 | # taggining it with vectorized=True |
---|
| 90 | if not getattr(kernel, 'vectorized', False): |
---|
| 91 | if dim == '2d': |
---|
| 92 | def vector_kernel(qx, qy, *args): |
---|
[eafc9fa] | 93 | """ |
---|
| 94 | Vectorized 2D kernel. |
---|
| 95 | """ |
---|
[3c56da87] | 96 | return np.array([kernel(qxi, qyi, *args) |
---|
| 97 | for qxi, qyi in zip(qx, qy)]) |
---|
[b3f6bc3] | 98 | else: |
---|
| 99 | def vector_kernel(q, *args): |
---|
[eafc9fa] | 100 | """ |
---|
| 101 | Vectorized 1D kernel. |
---|
| 102 | """ |
---|
[3c56da87] | 103 | return np.array([kernel(qi, *args) |
---|
| 104 | for qi in q]) |
---|
[b3f6bc3] | 105 | self.kernel = vector_kernel |
---|
| 106 | else: |
---|
| 107 | self.kernel = kernel |
---|
[17bbadd] | 108 | fixed_pars = model_info['partype']['fixed-' + dim] |
---|
| 109 | pd_pars = model_info['partype']['pd-' + dim] |
---|
| 110 | vol_pars = model_info['partype']['volume'] |
---|
[b3f6bc3] | 111 | |
---|
| 112 | # First two fixed pars are scale and background |
---|
[fcd7bbd] | 113 | pars = [p.name for p in model_info['parameters'][2:]] |
---|
[3c56da87] | 114 | offset = len(self.q_input.q_vectors) |
---|
| 115 | self.args = self.q_input.q_vectors + [None] * len(pars) |
---|
| 116 | self.fixed_index = np.array([pars.index(p) + offset |
---|
| 117 | for p in fixed_pars[2:]]) |
---|
| 118 | self.pd_index = np.array([pars.index(p) + offset |
---|
| 119 | for p in pd_pars]) |
---|
| 120 | self.vol_index = np.array([pars.index(p) + offset |
---|
| 121 | for p in vol_pars]) |
---|
[c85db69] | 122 | try: self.theta_index = pars.index('theta') + offset |
---|
[b3f6bc3] | 123 | except ValueError: self.theta_index = -1 |
---|
| 124 | |
---|
| 125 | # Caller needs fixed_pars and pd_pars |
---|
| 126 | self.fixed_pars = fixed_pars |
---|
| 127 | self.pd_pars = pd_pars |
---|
| 128 | |
---|
[6edb74a] | 129 | def __call__(self, fixed, pd, cutoff=1e-5): |
---|
[35647ab] | 130 | print("fixed",fixed) |
---|
| 131 | print("pd", pd) |
---|
[b3f6bc3] | 132 | args = self.args[:] # grab a copy of the args |
---|
| 133 | form, form_volume = self.kernel, self.info['form_volume'] |
---|
| 134 | # First two fixed |
---|
| 135 | scale, background = fixed[:2] |
---|
[c85db69] | 136 | for index, value in zip(self.fixed_index, fixed[2:]): |
---|
[b3f6bc3] | 137 | args[index] = float(value) |
---|
[c85db69] | 138 | res = _loops(form, form_volume, cutoff, scale, background, args, |
---|
[b3f6bc3] | 139 | pd, self.pd_index, self.vol_index, self.theta_index) |
---|
| 140 | |
---|
| 141 | return res |
---|
| 142 | |
---|
| 143 | def release(self): |
---|
[eafc9fa] | 144 | """ |
---|
| 145 | Free resources associated with the kernel. |
---|
| 146 | """ |
---|
[3c56da87] | 147 | self.q_input = None |
---|
[b3f6bc3] | 148 | |
---|
| 149 | def _loops(form, form_volume, cutoff, scale, background, |
---|
| 150 | args, pd, pd_index, vol_index, theta_index): |
---|
| 151 | """ |
---|
| 152 | *form* is the name of the form function, which should be vectorized over |
---|
| 153 | q, but otherwise have an interface like the opencl kernels, with the |
---|
| 154 | q parameters first followed by the individual parameters in the order |
---|
| 155 | given in model.parameters (see :mod:`sasmodels.generate`). |
---|
| 156 | |
---|
| 157 | *form_volume* calculates the volume of the shape. *vol_index* gives |
---|
| 158 | the list of volume parameters |
---|
| 159 | |
---|
| 160 | *cutoff* ignores the corners of the dispersion hypercube |
---|
| 161 | |
---|
| 162 | *scale*, *background* multiplies the resulting form and adds an offset |
---|
| 163 | |
---|
| 164 | *args* is the prepopulated set of arguments to the form function, starting |
---|
| 165 | with the q vectors, and including slots for all the parameters. The |
---|
| 166 | values for the parameters will be substituted with values from the |
---|
| 167 | dispersion functions. |
---|
| 168 | |
---|
| 169 | *pd* is the list of dispersion parameters |
---|
| 170 | |
---|
| 171 | *pd_index* are the indices of the dispersion parameters in *args* |
---|
| 172 | |
---|
| 173 | *vol_index* are the indices of the volume parameters in *args* |
---|
| 174 | |
---|
| 175 | *theta_index* is the index of the theta parameter for the sasview |
---|
| 176 | spherical correction, or -1 if there is no angular dispersion |
---|
| 177 | """ |
---|
| 178 | |
---|
[f734e7d] | 179 | ################################################################ |
---|
| 180 | # # |
---|
| 181 | # !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! # |
---|
| 182 | # !! !! # |
---|
| 183 | # !! KEEP THIS CODE CONSISTENT WITH KERNEL_TEMPLATE.C !! # |
---|
| 184 | # !! !! # |
---|
| 185 | # !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! # |
---|
| 186 | # # |
---|
| 187 | ################################################################ |
---|
[b3f6bc3] | 188 | |
---|
[35647ab] | 189 | #TODO: Wojtek's notes |
---|
| 190 | #TODO: The goal is to restructure polydispersity loop |
---|
| 191 | #so it allows fitting arbitrary polydispersity parameters |
---|
| 192 | #and it accepts cases like coupled parameters |
---|
[b3f6bc3] | 193 | weight = np.empty(len(pd), 'd') |
---|
[6edb74a] | 194 | if weight.size > 0: |
---|
| 195 | # weight vector, to be populated by polydispersity loops |
---|
| 196 | |
---|
| 197 | # identify which pd parameters are volume parameters |
---|
| 198 | vol_weight_index = np.array([(index in vol_index) for index in pd_index]) |
---|
| 199 | |
---|
| 200 | # Sort parameters in decreasing order of pd length |
---|
| 201 | Npd = np.array([len(pdi[0]) for pdi in pd], 'i') |
---|
| 202 | order = np.argsort(Npd)[::-1] |
---|
| 203 | stride = np.cumprod(Npd[order]) |
---|
| 204 | pd = [pd[index] for index in order] |
---|
| 205 | pd_index = pd_index[order] |
---|
| 206 | vol_weight_index = vol_weight_index[order] |
---|
| 207 | |
---|
| 208 | fast_value = pd[0][0] |
---|
| 209 | fast_weight = pd[0][1] |
---|
| 210 | else: |
---|
| 211 | stride = np.array([1]) |
---|
| 212 | vol_weight_index = slice(None, None) |
---|
[f734e7d] | 213 | # keep lint happy |
---|
| 214 | fast_value = [None] |
---|
| 215 | fast_weight = [None] |
---|
[b3f6bc3] | 216 | |
---|
| 217 | ret = np.zeros_like(args[0]) |
---|
| 218 | norm = np.zeros_like(ret) |
---|
| 219 | vol = np.zeros_like(ret) |
---|
| 220 | vol_norm = np.zeros_like(ret) |
---|
| 221 | for k in range(stride[-1]): |
---|
| 222 | # update polydispersity parameter values |
---|
[c85db69] | 223 | fast_index = k % stride[0] |
---|
[b3f6bc3] | 224 | if fast_index == 0: # bottom loop complete ... check all other loops |
---|
[6edb74a] | 225 | if weight.size > 0: |
---|
[c85db69] | 226 | for i, index, in enumerate(k % stride): |
---|
[6edb74a] | 227 | args[pd_index[i]] = pd[i][0][index] |
---|
| 228 | weight[i] = pd[i][1][index] |
---|
[b3f6bc3] | 229 | else: |
---|
| 230 | args[pd_index[0]] = fast_value[fast_index] |
---|
| 231 | weight[0] = fast_weight[fast_index] |
---|
[4c2c535] | 232 | |
---|
| 233 | # Computes the weight, and if it is not sufficient then ignore this |
---|
| 234 | # parameter set. |
---|
| 235 | # Note: could precompute w1*...*wn so we only need to multiply by w0 |
---|
[b3f6bc3] | 236 | w = np.prod(weight) |
---|
| 237 | if w > cutoff: |
---|
[3c56da87] | 238 | # Note: can precompute spherical correction if theta_index is not |
---|
| 239 | # the fast index. Correction factor for spherical integration |
---|
| 240 | #spherical_correction = abs(cos(pi*args[phi_index])) if phi_index>=0 else 1.0 |
---|
| 241 | spherical_correction = (abs(cos(pi * args[theta_index])) * pi / 2 |
---|
| 242 | if theta_index >= 0 else 1.0) |
---|
[f734e7d] | 243 | #spherical_correction = 1.0 |
---|
[4c2c535] | 244 | |
---|
| 245 | # Call the scattering function and adds it to the total. |
---|
| 246 | I = form(*args) |
---|
| 247 | if np.isnan(I).any(): continue |
---|
| 248 | ret += w * I * spherical_correction |
---|
| 249 | norm += w |
---|
[b3f6bc3] | 250 | |
---|
| 251 | # Volume normalization. |
---|
[3c56da87] | 252 | # If there are "volume" polydispersity parameters, then these |
---|
| 253 | # will be used to call the form_volume function from the user |
---|
| 254 | # supplied kernel, and accumulate a normalized weight. |
---|
| 255 | # Note: can precompute volume norm if fast index is not a volume |
---|
[b3f6bc3] | 256 | if form_volume: |
---|
| 257 | vol_args = [args[index] for index in vol_index] |
---|
| 258 | vol_weight = np.prod(weight[vol_weight_index]) |
---|
[4c2c535] | 259 | vol += vol_weight * form_volume(*vol_args) |
---|
| 260 | vol_norm += vol_weight |
---|
[b3f6bc3] | 261 | |
---|
[c85db69] | 262 | positive = (vol * vol_norm != 0.0) |
---|
[b3f6bc3] | 263 | ret[positive] *= vol_norm[positive] / vol[positive] |
---|
[c85db69] | 264 | result = scale * ret / norm + background |
---|
[b3f6bc3] | 265 | return result |
---|
[352964b] | 266 | ======= |
---|
[a84a0ca] | 267 | """ |
---|
| 268 | Python driver for python kernels |
---|
| 269 | |
---|
| 270 | Calls the kernel with a vector of $q$ values for a single parameter set. |
---|
| 271 | Polydispersity is supported by looping over different parameter sets and |
---|
| 272 | summing the results. The interface to :class:`PyModel` matches those for |
---|
| 273 | :class:`kernelcl.GpuModel` and :class:`kerneldll.DllModel`. |
---|
| 274 | """ |
---|
| 275 | import numpy as np |
---|
| 276 | from numpy import pi, cos |
---|
| 277 | |
---|
| 278 | from .generate import F64 |
---|
| 279 | |
---|
| 280 | class PyModel(object): |
---|
| 281 | """ |
---|
| 282 | Wrapper for pure python models. |
---|
| 283 | """ |
---|
| 284 | def __init__(self, model_info): |
---|
| 285 | self.info = model_info |
---|
| 286 | |
---|
| 287 | def __call__(self, q_vectors): |
---|
| 288 | q_input = PyInput(q_vectors, dtype=F64) |
---|
| 289 | kernel = self.info['Iqxy'] if q_input.is_2d else self.info['Iq'] |
---|
| 290 | return PyKernel(kernel, self.info, q_input) |
---|
| 291 | |
---|
| 292 | def release(self): |
---|
| 293 | """ |
---|
| 294 | Free resources associated with the model. |
---|
| 295 | """ |
---|
| 296 | pass |
---|
| 297 | |
---|
| 298 | class PyInput(object): |
---|
| 299 | """ |
---|
| 300 | Make q data available to the gpu. |
---|
| 301 | |
---|
| 302 | *q_vectors* is a list of q vectors, which will be *[q]* for 1-D data, |
---|
| 303 | and *[qx, qy]* for 2-D data. Internally, the vectors will be reallocated |
---|
| 304 | to get the best performance on OpenCL, which may involve shifting and |
---|
| 305 | stretching the array to better match the memory architecture. Additional |
---|
| 306 | points will be evaluated with *q=1e-3*. |
---|
| 307 | |
---|
| 308 | *dtype* is the data type for the q vectors. The data type should be |
---|
| 309 | set to match that of the kernel, which is an attribute of |
---|
| 310 | :class:`GpuProgram`. Note that not all kernels support double |
---|
| 311 | precision, so even if the program was created for double precision, |
---|
| 312 | the *GpuProgram.dtype* may be single precision. |
---|
| 313 | |
---|
| 314 | Call :meth:`release` when complete. Even if not called directly, the |
---|
| 315 | buffer will be released when the data object is freed. |
---|
| 316 | """ |
---|
| 317 | def __init__(self, q_vectors, dtype): |
---|
| 318 | self.nq = q_vectors[0].size |
---|
| 319 | self.dtype = dtype |
---|
| 320 | self.is_2d = (len(q_vectors) == 2) |
---|
| 321 | self.q_vectors = [np.ascontiguousarray(q, self.dtype) for q in q_vectors] |
---|
| 322 | self.q_pointers = [q.ctypes.data for q in self.q_vectors] |
---|
| 323 | |
---|
| 324 | def release(self): |
---|
| 325 | """ |
---|
| 326 | Free resources associated with the model inputs. |
---|
| 327 | """ |
---|
| 328 | self.q_vectors = [] |
---|
| 329 | |
---|
| 330 | class PyKernel(object): |
---|
| 331 | """ |
---|
| 332 | Callable SAS kernel. |
---|
| 333 | |
---|
| 334 | *kernel* is the DllKernel object to call. |
---|
| 335 | |
---|
| 336 | *model_info* is the module information |
---|
| 337 | |
---|
| 338 | *q_input* is the DllInput q vectors at which the kernel should be |
---|
| 339 | evaluated. |
---|
| 340 | |
---|
| 341 | The resulting call method takes the *pars*, a list of values for |
---|
| 342 | the fixed parameters to the kernel, and *pd_pars*, a list of (value,weight) |
---|
| 343 | vectors for the polydisperse parameters. *cutoff* determines the |
---|
| 344 | integration limits: any points with combined weight less than *cutoff* |
---|
| 345 | will not be calculated. |
---|
| 346 | |
---|
| 347 | Call :meth:`release` when done with the kernel instance. |
---|
| 348 | """ |
---|
| 349 | def __init__(self, kernel, model_info, q_input): |
---|
| 350 | self.info = model_info |
---|
| 351 | self.q_input = q_input |
---|
| 352 | self.res = np.empty(q_input.nq, q_input.dtype) |
---|
| 353 | dim = '2d' if q_input.is_2d else '1d' |
---|
| 354 | # Loop over q unless user promises that the kernel is vectorized by |
---|
| 355 | # taggining it with vectorized=True |
---|
| 356 | if not getattr(kernel, 'vectorized', False): |
---|
| 357 | if dim == '2d': |
---|
| 358 | def vector_kernel(qx, qy, *args): |
---|
| 359 | """ |
---|
| 360 | Vectorized 2D kernel. |
---|
| 361 | """ |
---|
| 362 | return np.array([kernel(qxi, qyi, *args) |
---|
| 363 | for qxi, qyi in zip(qx, qy)]) |
---|
| 364 | else: |
---|
| 365 | def vector_kernel(q, *args): |
---|
| 366 | """ |
---|
| 367 | Vectorized 1D kernel. |
---|
| 368 | """ |
---|
| 369 | return np.array([kernel(qi, *args) |
---|
| 370 | for qi in q]) |
---|
| 371 | self.kernel = vector_kernel |
---|
| 372 | else: |
---|
| 373 | self.kernel = kernel |
---|
| 374 | fixed_pars = model_info['partype']['fixed-' + dim] |
---|
| 375 | pd_pars = model_info['partype']['pd-' + dim] |
---|
| 376 | vol_pars = model_info['partype']['volume'] |
---|
| 377 | |
---|
| 378 | # First two fixed pars are scale and background |
---|
| 379 | pars = [p.name for p in model_info['parameters'][2:]] |
---|
| 380 | offset = len(self.q_input.q_vectors) |
---|
| 381 | self.args = self.q_input.q_vectors + [None] * len(pars) |
---|
| 382 | self.fixed_index = np.array([pars.index(p) + offset |
---|
| 383 | for p in fixed_pars[2:]]) |
---|
| 384 | self.pd_index = np.array([pars.index(p) + offset |
---|
| 385 | for p in pd_pars]) |
---|
| 386 | self.vol_index = np.array([pars.index(p) + offset |
---|
| 387 | for p in vol_pars]) |
---|
| 388 | try: self.theta_index = pars.index('theta') + offset |
---|
| 389 | except ValueError: self.theta_index = -1 |
---|
| 390 | |
---|
| 391 | # Caller needs fixed_pars and pd_pars |
---|
| 392 | self.fixed_pars = fixed_pars |
---|
| 393 | self.pd_pars = pd_pars |
---|
| 394 | |
---|
| 395 | def __call__(self, fixed, pd, cutoff=1e-5): |
---|
| 396 | #print("fixed",fixed) |
---|
| 397 | #print("pd", pd) |
---|
| 398 | args = self.args[:] # grab a copy of the args |
---|
| 399 | form, form_volume = self.kernel, self.info['form_volume'] |
---|
| 400 | # First two fixed |
---|
| 401 | scale, background = fixed[:2] |
---|
| 402 | for index, value in zip(self.fixed_index, fixed[2:]): |
---|
| 403 | args[index] = float(value) |
---|
| 404 | res = _loops(form, form_volume, cutoff, scale, background, args, |
---|
| 405 | pd, self.pd_index, self.vol_index, self.theta_index) |
---|
| 406 | |
---|
| 407 | return res |
---|
| 408 | |
---|
| 409 | def release(self): |
---|
| 410 | """ |
---|
| 411 | Free resources associated with the kernel. |
---|
| 412 | """ |
---|
| 413 | self.q_input = None |
---|
| 414 | |
---|
| 415 | def _loops(form, form_volume, cutoff, scale, background, |
---|
| 416 | args, pd, pd_index, vol_index, theta_index): |
---|
| 417 | """ |
---|
| 418 | *form* is the name of the form function, which should be vectorized over |
---|
| 419 | q, but otherwise have an interface like the opencl kernels, with the |
---|
| 420 | q parameters first followed by the individual parameters in the order |
---|
| 421 | given in model.parameters (see :mod:`sasmodels.generate`). |
---|
| 422 | |
---|
| 423 | *form_volume* calculates the volume of the shape. *vol_index* gives |
---|
| 424 | the list of volume parameters |
---|
| 425 | |
---|
| 426 | *cutoff* ignores the corners of the dispersion hypercube |
---|
| 427 | |
---|
| 428 | *scale*, *background* multiplies the resulting form and adds an offset |
---|
| 429 | |
---|
| 430 | *args* is the prepopulated set of arguments to the form function, starting |
---|
| 431 | with the q vectors, and including slots for all the parameters. The |
---|
| 432 | values for the parameters will be substituted with values from the |
---|
| 433 | dispersion functions. |
---|
| 434 | |
---|
| 435 | *pd* is the list of dispersion parameters |
---|
| 436 | |
---|
| 437 | *pd_index* are the indices of the dispersion parameters in *args* |
---|
| 438 | |
---|
| 439 | *vol_index* are the indices of the volume parameters in *args* |
---|
| 440 | |
---|
| 441 | *theta_index* is the index of the theta parameter for the sasview |
---|
| 442 | spherical correction, or -1 if there is no angular dispersion |
---|
| 443 | """ |
---|
| 444 | |
---|
| 445 | ################################################################ |
---|
| 446 | # # |
---|
| 447 | # !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! # |
---|
| 448 | # !! !! # |
---|
| 449 | # !! KEEP THIS CODE CONSISTENT WITH KERNEL_TEMPLATE.C !! # |
---|
| 450 | # !! !! # |
---|
| 451 | # !!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!! # |
---|
| 452 | # # |
---|
| 453 | ################################################################ |
---|
| 454 | |
---|
| 455 | weight = np.empty(len(pd), 'd') |
---|
| 456 | if weight.size > 0: |
---|
| 457 | # weight vector, to be populated by polydispersity loops |
---|
| 458 | |
---|
| 459 | # identify which pd parameters are volume parameters |
---|
| 460 | vol_weight_index = np.array([(index in vol_index) for index in pd_index]) |
---|
| 461 | |
---|
| 462 | # Sort parameters in decreasing order of pd length |
---|
| 463 | Npd = np.array([len(pdi[0]) for pdi in pd], 'i') |
---|
| 464 | order = np.argsort(Npd)[::-1] |
---|
| 465 | stride = np.cumprod(Npd[order]) |
---|
| 466 | pd = [pd[index] for index in order] |
---|
| 467 | pd_index = pd_index[order] |
---|
| 468 | vol_weight_index = vol_weight_index[order] |
---|
| 469 | |
---|
| 470 | fast_value = pd[0][0] |
---|
| 471 | fast_weight = pd[0][1] |
---|
| 472 | else: |
---|
| 473 | stride = np.array([1]) |
---|
| 474 | vol_weight_index = slice(None, None) |
---|
| 475 | # keep lint happy |
---|
| 476 | fast_value = [None] |
---|
| 477 | fast_weight = [None] |
---|
| 478 | |
---|
| 479 | ret = np.zeros_like(args[0]) |
---|
| 480 | norm = np.zeros_like(ret) |
---|
| 481 | vol = np.zeros_like(ret) |
---|
| 482 | vol_norm = np.zeros_like(ret) |
---|
| 483 | for k in range(stride[-1]): |
---|
| 484 | # update polydispersity parameter values |
---|
| 485 | fast_index = k % stride[0] |
---|
| 486 | if fast_index == 0: # bottom loop complete ... check all other loops |
---|
| 487 | if weight.size > 0: |
---|
| 488 | for i, index, in enumerate(k % stride): |
---|
| 489 | args[pd_index[i]] = pd[i][0][index] |
---|
| 490 | weight[i] = pd[i][1][index] |
---|
| 491 | else: |
---|
| 492 | args[pd_index[0]] = fast_value[fast_index] |
---|
| 493 | weight[0] = fast_weight[fast_index] |
---|
| 494 | |
---|
| 495 | # Computes the weight, and if it is not sufficient then ignore this |
---|
| 496 | # parameter set. |
---|
| 497 | # Note: could precompute w1*...*wn so we only need to multiply by w0 |
---|
| 498 | w = np.prod(weight) |
---|
| 499 | if w > cutoff: |
---|
| 500 | # Note: can precompute spherical correction if theta_index is not |
---|
| 501 | # the fast index. Correction factor for spherical integration |
---|
| 502 | #spherical_correction = abs(cos(pi*args[phi_index])) if phi_index>=0 else 1.0 |
---|
| 503 | spherical_correction = (abs(cos(pi * args[theta_index])) * pi / 2 |
---|
| 504 | if theta_index >= 0 else 1.0) |
---|
| 505 | #spherical_correction = 1.0 |
---|
| 506 | |
---|
| 507 | # Call the scattering function and adds it to the total. |
---|
| 508 | I = form(*args) |
---|
| 509 | if np.isnan(I).any(): continue |
---|
| 510 | ret += w * I * spherical_correction |
---|
| 511 | norm += w |
---|
| 512 | |
---|
| 513 | # Volume normalization. |
---|
| 514 | # If there are "volume" polydispersity parameters, then these |
---|
| 515 | # will be used to call the form_volume function from the user |
---|
| 516 | # supplied kernel, and accumulate a normalized weight. |
---|
| 517 | # Note: can precompute volume norm if fast index is not a volume |
---|
| 518 | if form_volume: |
---|
| 519 | vol_args = [args[index] for index in vol_index] |
---|
| 520 | vol_weight = np.prod(weight[vol_weight_index]) |
---|
| 521 | vol += vol_weight * form_volume(*vol_args) |
---|
| 522 | vol_norm += vol_weight |
---|
| 523 | |
---|
| 524 | positive = (vol * vol_norm != 0.0) |
---|
| 525 | ret[positive] *= vol_norm[positive] / vol[positive] |
---|
| 526 | result = scale * ret / norm + background |
---|
| 527 | return result |
---|