[87985ca] | 1 | """ |
---|
| 2 | Sasview model constructor. |
---|
| 3 | |
---|
| 4 | Given a module defining an OpenCL kernel such as sasmodels.models.cylinder, |
---|
| 5 | create a sasview model class to run that kernel as follows:: |
---|
| 6 | |
---|
[92d38285] | 7 | from sasmodels.sasview_model import load_custom_model |
---|
| 8 | CylinderModel = load_custom_model('sasmodels/models/cylinder.py') |
---|
[87985ca] | 9 | """ |
---|
[4d76711] | 10 | from __future__ import print_function |
---|
[87985ca] | 11 | |
---|
[ce27e21] | 12 | import math |
---|
| 13 | from copy import deepcopy |
---|
[2622b3f] | 14 | import collections |
---|
[4d76711] | 15 | import traceback |
---|
| 16 | import logging |
---|
[724257c] | 17 | from os.path import basename, splitext, abspath, getmtime |
---|
[9f8ade1] | 18 | try: |
---|
| 19 | import _thread as thread |
---|
| 20 | except ImportError: |
---|
| 21 | import thread |
---|
[ce27e21] | 22 | |
---|
[7ae2b7f] | 23 | import numpy as np # type: ignore |
---|
[ce27e21] | 24 | |
---|
[aa4946b] | 25 | from . import core |
---|
[4d76711] | 26 | from . import custom |
---|
[a80e64c] | 27 | from . import product |
---|
[72a081d] | 28 | from . import generate |
---|
[fb5914f] | 29 | from . import weights |
---|
[6d6508e] | 30 | from . import modelinfo |
---|
[bde38b5] | 31 | from .details import make_kernel_args, dispersion_mesh |
---|
[ff7119b] | 32 | |
---|
[fa5fd8d] | 33 | try: |
---|
[60f03de] | 34 | from typing import Dict, Mapping, Any, Sequence, Tuple, NamedTuple, List, Optional, Union, Callable |
---|
[fa5fd8d] | 35 | from .modelinfo import ModelInfo, Parameter |
---|
| 36 | from .kernel import KernelModel |
---|
| 37 | MultiplicityInfoType = NamedTuple( |
---|
| 38 | 'MuliplicityInfo', |
---|
| 39 | [("number", int), ("control", str), ("choices", List[str]), |
---|
| 40 | ("x_axis_label", str)]) |
---|
[60f03de] | 41 | SasviewModelType = Callable[[int], "SasviewModel"] |
---|
[fa5fd8d] | 42 | except ImportError: |
---|
| 43 | pass |
---|
| 44 | |
---|
[724257c] | 45 | logger = logging.getLogger(__name__) |
---|
| 46 | |
---|
[a38b065] | 47 | calculation_lock = thread.allocate_lock() |
---|
| 48 | |
---|
[724257c] | 49 | #: True if pre-existing plugins, with the old names and parameters, should |
---|
| 50 | #: continue to be supported. |
---|
[c95dfc63] | 51 | SUPPORT_OLD_STYLE_PLUGINS = True |
---|
| 52 | |
---|
[fa5fd8d] | 53 | # TODO: separate x_axis_label from multiplicity info |
---|
| 54 | MultiplicityInfo = collections.namedtuple( |
---|
| 55 | 'MultiplicityInfo', |
---|
| 56 | ["number", "control", "choices", "x_axis_label"], |
---|
| 57 | ) |
---|
| 58 | |
---|
[724257c] | 59 | #: set of defined models (standard and custom) |
---|
| 60 | MODELS = {} # type: Dict[str, SasviewModelType] |
---|
| 61 | #: custom model {path: model} mapping so we can check timestamps |
---|
| 62 | MODEL_BY_PATH = {} # type: Dict[str, SasviewModelType] |
---|
| 63 | |
---|
[92d38285] | 64 | def find_model(modelname): |
---|
[b32dafd] | 65 | # type: (str) -> SasviewModelType |
---|
| 66 | """ |
---|
| 67 | Find a model by name. If the model name ends in py, try loading it from |
---|
| 68 | custom models, otherwise look for it in the list of builtin models. |
---|
| 69 | """ |
---|
[92d38285] | 70 | # TODO: used by sum/product model to load an existing model |
---|
| 71 | # TODO: doesn't handle custom models properly |
---|
| 72 | if modelname.endswith('.py'): |
---|
| 73 | return load_custom_model(modelname) |
---|
| 74 | elif modelname in MODELS: |
---|
| 75 | return MODELS[modelname] |
---|
| 76 | else: |
---|
| 77 | raise ValueError("unknown model %r"%modelname) |
---|
| 78 | |
---|
[56b2687] | 79 | |
---|
[fa5fd8d] | 80 | # TODO: figure out how to say that the return type is a subclass |
---|
[4d76711] | 81 | def load_standard_models(): |
---|
[60f03de] | 82 | # type: () -> List[SasviewModelType] |
---|
[4d76711] | 83 | """ |
---|
| 84 | Load and return the list of predefined models. |
---|
| 85 | |
---|
| 86 | If there is an error loading a model, then a traceback is logged and the |
---|
| 87 | model is not returned. |
---|
| 88 | """ |
---|
| 89 | for name in core.list_models(): |
---|
| 90 | try: |
---|
[92d38285] | 91 | MODELS[name] = _make_standard_model(name) |
---|
[ee8f734] | 92 | except Exception: |
---|
[724257c] | 93 | logger.error(traceback.format_exc()) |
---|
[c95dfc63] | 94 | if SUPPORT_OLD_STYLE_PLUGINS: |
---|
| 95 | _register_old_models() |
---|
| 96 | |
---|
[724257c] | 97 | return list(MODELS.values()) |
---|
[de97440] | 98 | |
---|
[4d76711] | 99 | |
---|
| 100 | def load_custom_model(path): |
---|
[60f03de] | 101 | # type: (str) -> SasviewModelType |
---|
[4d76711] | 102 | """ |
---|
| 103 | Load a custom model given the model path. |
---|
[ff7119b] | 104 | """ |
---|
[724257c] | 105 | model = MODEL_BY_PATH.get(path, None) |
---|
| 106 | if model is not None and model.timestamp == getmtime(path): |
---|
| 107 | #logger.info("Model already loaded %s", path) |
---|
| 108 | return model |
---|
| 109 | |
---|
| 110 | #logger.info("Loading model %s", path) |
---|
[4d76711] | 111 | kernel_module = custom.load_custom_kernel_module(path) |
---|
[724257c] | 112 | if hasattr(kernel_module, 'Model'): |
---|
[92d38285] | 113 | model = kernel_module.Model |
---|
[9457498] | 114 | # Old style models do not set the name in the class attributes, so |
---|
| 115 | # set it here; this name will be overridden when the object is created |
---|
| 116 | # with an instance variable that has the same value. |
---|
| 117 | if model.name == "": |
---|
| 118 | model.name = splitext(basename(path))[0] |
---|
[20a70bc] | 119 | if not hasattr(model, 'filename'): |
---|
[724257c] | 120 | model.filename = abspath(kernel_module.__file__).replace('.pyc', '.py') |
---|
[e4bf271] | 121 | if not hasattr(model, 'id'): |
---|
| 122 | model.id = splitext(basename(model.filename))[0] |
---|
[724257c] | 123 | else: |
---|
[56b2687] | 124 | model_info = modelinfo.make_model_info(kernel_module) |
---|
[bcdd6c9] | 125 | model = make_model_from_info(model_info) |
---|
[724257c] | 126 | model.timestamp = getmtime(path) |
---|
[ed10b57] | 127 | |
---|
[2f2c70c] | 128 | # If a model name already exists and we are loading a different model, |
---|
| 129 | # use the model file name as the model name. |
---|
| 130 | if model.name in MODELS and not model.filename == MODELS[model.name].filename: |
---|
| 131 | _previous_name = model.name |
---|
| 132 | model.name = model.id |
---|
[bf8c271] | 133 | |
---|
[2f2c70c] | 134 | # If the new model name is still in the model list (for instance, |
---|
| 135 | # if we put a cylinder.py in our plug-in directory), then append |
---|
| 136 | # an identifier. |
---|
| 137 | if model.name in MODELS and not model.filename == MODELS[model.name].filename: |
---|
| 138 | model.name = model.id + '_user' |
---|
[724257c] | 139 | logger.info("Model %s already exists: using %s [%s]", |
---|
| 140 | _previous_name, model.name, model.filename) |
---|
[ed10b57] | 141 | |
---|
[92d38285] | 142 | MODELS[model.name] = model |
---|
[724257c] | 143 | MODEL_BY_PATH[path] = model |
---|
[92d38285] | 144 | return model |
---|
[4d76711] | 145 | |
---|
[87985ca] | 146 | |
---|
[bcdd6c9] | 147 | def make_model_from_info(model_info): |
---|
| 148 | # type: (ModelInfo) -> SasviewModelType |
---|
| 149 | """ |
---|
| 150 | Convert *model_info* into a SasView model wrapper. |
---|
| 151 | """ |
---|
| 152 | def __init__(self, multiplicity=None): |
---|
| 153 | SasviewModel.__init__(self, multiplicity=multiplicity) |
---|
| 154 | attrs = _generate_model_attributes(model_info) |
---|
| 155 | attrs['__init__'] = __init__ |
---|
| 156 | attrs['filename'] = model_info.filename |
---|
| 157 | ConstructedModel = type(model_info.name, (SasviewModel,), attrs) # type: SasviewModelType |
---|
| 158 | return ConstructedModel |
---|
| 159 | |
---|
| 160 | |
---|
[4d76711] | 161 | def _make_standard_model(name): |
---|
[60f03de] | 162 | # type: (str) -> SasviewModelType |
---|
[ff7119b] | 163 | """ |
---|
[4d76711] | 164 | Load the sasview model defined by *name*. |
---|
[72a081d] | 165 | |
---|
[4d76711] | 166 | *name* can be a standard model name or a path to a custom model. |
---|
[87985ca] | 167 | |
---|
[4d76711] | 168 | Returns a class that can be used directly as a sasview model. |
---|
[ff7119b] | 169 | """ |
---|
[4d76711] | 170 | kernel_module = generate.load_kernel_module(name) |
---|
[fa5fd8d] | 171 | model_info = modelinfo.make_model_info(kernel_module) |
---|
[bcdd6c9] | 172 | return make_model_from_info(model_info) |
---|
[72a081d] | 173 | |
---|
| 174 | |
---|
[724257c] | 175 | def _register_old_models(): |
---|
| 176 | # type: () -> None |
---|
| 177 | """ |
---|
| 178 | Place the new models into sasview under the old names. |
---|
| 179 | |
---|
| 180 | Monkey patch sas.sascalc.fit as sas.models so that sas.models.pluginmodel |
---|
| 181 | is available to the plugin modules. |
---|
| 182 | """ |
---|
| 183 | import sys |
---|
| 184 | import sas # needed in order to set sas.models |
---|
| 185 | import sas.sascalc.fit |
---|
| 186 | sys.modules['sas.models'] = sas.sascalc.fit |
---|
| 187 | sas.models = sas.sascalc.fit |
---|
| 188 | |
---|
| 189 | import sas.models |
---|
| 190 | from sasmodels.conversion_table import CONVERSION_TABLE |
---|
| 191 | for new_name, conversion in CONVERSION_TABLE.get((3, 1, 2), {}).items(): |
---|
| 192 | # CoreShellEllipsoidModel => core_shell_ellipsoid:1 |
---|
| 193 | new_name = new_name.split(':')[0] |
---|
| 194 | old_name = conversion[0] if len(conversion) < 3 else conversion[2] |
---|
| 195 | module_attrs = {old_name: find_model(new_name)} |
---|
| 196 | ConstructedModule = type(old_name, (), module_attrs) |
---|
| 197 | old_path = 'sas.models.' + old_name |
---|
| 198 | setattr(sas.models, old_path, ConstructedModule) |
---|
| 199 | sys.modules[old_path] = ConstructedModule |
---|
| 200 | |
---|
| 201 | |
---|
[a80e64c] | 202 | def MultiplicationModel(form_factor, structure_factor): |
---|
| 203 | # type: ("SasviewModel", "SasviewModel") -> "SasviewModel" |
---|
| 204 | model_info = product.make_product_info(form_factor._model_info, |
---|
| 205 | structure_factor._model_info) |
---|
[bcdd6c9] | 206 | ConstructedModel = make_model_from_info(model_info) |
---|
[edb0f85] | 207 | return ConstructedModel(form_factor.multiplicity) |
---|
[a80e64c] | 208 | |
---|
[ce27e21] | 209 | |
---|
[fa5fd8d] | 210 | def _generate_model_attributes(model_info): |
---|
| 211 | # type: (ModelInfo) -> Dict[str, Any] |
---|
| 212 | """ |
---|
| 213 | Generate the class attributes for the model. |
---|
| 214 | |
---|
| 215 | This should include all the information necessary to query the model |
---|
| 216 | details so that you do not need to instantiate a model to query it. |
---|
| 217 | |
---|
| 218 | All the attributes should be immutable to avoid accidents. |
---|
| 219 | """ |
---|
| 220 | |
---|
| 221 | # TODO: allow model to override axis labels input/output name/unit |
---|
| 222 | |
---|
[a18c5b3] | 223 | # Process multiplicity |
---|
[fa5fd8d] | 224 | non_fittable = [] # type: List[str] |
---|
[04045f4] | 225 | xlabel = model_info.profile_axes[0] if model_info.profile is not None else "" |
---|
| 226 | variants = MultiplicityInfo(0, "", [], xlabel) |
---|
[a18c5b3] | 227 | for p in model_info.parameters.kernel_parameters: |
---|
[04045f4] | 228 | if p.name == model_info.control: |
---|
[fa5fd8d] | 229 | non_fittable.append(p.name) |
---|
[04045f4] | 230 | variants = MultiplicityInfo( |
---|
[ce176ca] | 231 | len(p.choices) if p.choices else int(p.limits[1]), |
---|
| 232 | p.name, p.choices, xlabel |
---|
[fa5fd8d] | 233 | ) |
---|
| 234 | break |
---|
| 235 | |
---|
[50ec515] | 236 | # Only a single drop-down list parameter available |
---|
| 237 | fun_list = [] |
---|
| 238 | for p in model_info.parameters.kernel_parameters: |
---|
| 239 | if p.choices: |
---|
| 240 | fun_list = p.choices |
---|
| 241 | if p.length > 1: |
---|
| 242 | non_fittable.extend(p.id+str(k) for k in range(1, p.length+1)) |
---|
| 243 | break |
---|
| 244 | |
---|
[a18c5b3] | 245 | # Organize parameter sets |
---|
[fa5fd8d] | 246 | orientation_params = [] |
---|
| 247 | magnetic_params = [] |
---|
| 248 | fixed = [] |
---|
[85fe7f8] | 249 | for p in model_info.parameters.user_parameters({}, is2d=True): |
---|
[fa5fd8d] | 250 | if p.type == 'orientation': |
---|
| 251 | orientation_params.append(p.name) |
---|
| 252 | orientation_params.append(p.name+".width") |
---|
| 253 | fixed.append(p.name+".width") |
---|
[32e3c9b] | 254 | elif p.type == 'magnetic': |
---|
[fa5fd8d] | 255 | orientation_params.append(p.name) |
---|
| 256 | magnetic_params.append(p.name) |
---|
| 257 | fixed.append(p.name+".width") |
---|
[a18c5b3] | 258 | |
---|
[32e3c9b] | 259 | |
---|
[a18c5b3] | 260 | # Build class dictionary |
---|
| 261 | attrs = {} # type: Dict[str, Any] |
---|
| 262 | attrs['_model_info'] = model_info |
---|
| 263 | attrs['name'] = model_info.name |
---|
| 264 | attrs['id'] = model_info.id |
---|
| 265 | attrs['description'] = model_info.description |
---|
| 266 | attrs['category'] = model_info.category |
---|
| 267 | attrs['is_structure_factor'] = model_info.structure_factor |
---|
| 268 | attrs['is_form_factor'] = model_info.ER is not None |
---|
| 269 | attrs['is_multiplicity_model'] = variants[0] > 1 |
---|
| 270 | attrs['multiplicity_info'] = variants |
---|
[fa5fd8d] | 271 | attrs['orientation_params'] = tuple(orientation_params) |
---|
| 272 | attrs['magnetic_params'] = tuple(magnetic_params) |
---|
| 273 | attrs['fixed'] = tuple(fixed) |
---|
| 274 | attrs['non_fittable'] = tuple(non_fittable) |
---|
[50ec515] | 275 | attrs['fun_list'] = tuple(fun_list) |
---|
[fa5fd8d] | 276 | |
---|
| 277 | return attrs |
---|
[4d76711] | 278 | |
---|
[ce27e21] | 279 | class SasviewModel(object): |
---|
| 280 | """ |
---|
| 281 | Sasview wrapper for opencl/ctypes model. |
---|
| 282 | """ |
---|
[fa5fd8d] | 283 | # Model parameters for the specific model are set in the class constructor |
---|
| 284 | # via the _generate_model_attributes function, which subclasses |
---|
| 285 | # SasviewModel. They are included here for typing and documentation |
---|
| 286 | # purposes. |
---|
| 287 | _model = None # type: KernelModel |
---|
| 288 | _model_info = None # type: ModelInfo |
---|
| 289 | #: load/save name for the model |
---|
| 290 | id = None # type: str |
---|
| 291 | #: display name for the model |
---|
| 292 | name = None # type: str |
---|
| 293 | #: short model description |
---|
| 294 | description = None # type: str |
---|
| 295 | #: default model category |
---|
| 296 | category = None # type: str |
---|
| 297 | |
---|
| 298 | #: names of the orientation parameters in the order they appear |
---|
[724257c] | 299 | orientation_params = None # type: List[str] |
---|
[fa5fd8d] | 300 | #: names of the magnetic parameters in the order they appear |
---|
[724257c] | 301 | magnetic_params = None # type: List[str] |
---|
[fa5fd8d] | 302 | #: names of the fittable parameters |
---|
[724257c] | 303 | fixed = None # type: List[str] |
---|
[fa5fd8d] | 304 | # TODO: the attribute fixed is ill-named |
---|
| 305 | |
---|
| 306 | # Axis labels |
---|
| 307 | input_name = "Q" |
---|
| 308 | input_unit = "A^{-1}" |
---|
| 309 | output_name = "Intensity" |
---|
| 310 | output_unit = "cm^{-1}" |
---|
| 311 | |
---|
| 312 | #: default cutoff for polydispersity |
---|
| 313 | cutoff = 1e-5 |
---|
| 314 | |
---|
| 315 | # Note: Use non-mutable values for class attributes to avoid errors |
---|
| 316 | #: parameters that are not fitted |
---|
| 317 | non_fittable = () # type: Sequence[str] |
---|
| 318 | |
---|
| 319 | #: True if model should appear as a structure factor |
---|
| 320 | is_structure_factor = False |
---|
| 321 | #: True if model should appear as a form factor |
---|
| 322 | is_form_factor = False |
---|
| 323 | #: True if model has multiplicity |
---|
| 324 | is_multiplicity_model = False |
---|
[1f35235] | 325 | #: Multiplicity information |
---|
[fa5fd8d] | 326 | multiplicity_info = None # type: MultiplicityInfoType |
---|
| 327 | |
---|
| 328 | # Per-instance variables |
---|
| 329 | #: parameter {name: value} mapping |
---|
| 330 | params = None # type: Dict[str, float] |
---|
| 331 | #: values for dispersion width, npts, nsigmas and type |
---|
| 332 | dispersion = None # type: Dict[str, Any] |
---|
| 333 | #: units and limits for each parameter |
---|
[60f03de] | 334 | details = None # type: Dict[str, Sequence[Any]] |
---|
| 335 | # # actual type is Dict[str, List[str, float, float]] |
---|
[04dc697] | 336 | #: multiplicity value, or None if no multiplicity on the model |
---|
[fa5fd8d] | 337 | multiplicity = None # type: Optional[int] |
---|
[04dc697] | 338 | #: memory for polydispersity array if using ArrayDispersion (used by sasview). |
---|
| 339 | _persistency_dict = None # type: Dict[str, Tuple[np.ndarray, np.ndarray]] |
---|
[fa5fd8d] | 340 | |
---|
| 341 | def __init__(self, multiplicity=None): |
---|
[04dc697] | 342 | # type: (Optional[int]) -> None |
---|
[2622b3f] | 343 | |
---|
[04045f4] | 344 | # TODO: _persistency_dict to persistency_dict throughout sasview |
---|
| 345 | # TODO: refactor multiplicity to encompass variants |
---|
| 346 | # TODO: dispersion should be a class |
---|
[fa5fd8d] | 347 | # TODO: refactor multiplicity info |
---|
| 348 | # TODO: separate profile view from multiplicity |
---|
| 349 | # The button label, x and y axis labels and scale need to be under |
---|
| 350 | # the control of the model, not the fit page. Maximum flexibility, |
---|
| 351 | # the fit page would supply the canvas and the profile could plot |
---|
| 352 | # how it wants, but this assumes matplotlib. Next level is that |
---|
| 353 | # we provide some sort of data description including title, labels |
---|
| 354 | # and lines to plot. |
---|
| 355 | |
---|
[1f35235] | 356 | # Get the list of hidden parameters given the multiplicity |
---|
[04045f4] | 357 | # Don't include multiplicity in the list of parameters |
---|
[fa5fd8d] | 358 | self.multiplicity = multiplicity |
---|
[04045f4] | 359 | if multiplicity is not None: |
---|
| 360 | hidden = self._model_info.get_hidden_parameters(multiplicity) |
---|
| 361 | hidden |= set([self.multiplicity_info.control]) |
---|
| 362 | else: |
---|
| 363 | hidden = set() |
---|
[8f93522] | 364 | if self._model_info.structure_factor: |
---|
| 365 | hidden.add('scale') |
---|
| 366 | hidden.add('background') |
---|
| 367 | self._model_info.parameters.defaults['background'] = 0. |
---|
[04045f4] | 368 | |
---|
[04dc697] | 369 | self._persistency_dict = {} |
---|
[fa5fd8d] | 370 | self.params = collections.OrderedDict() |
---|
[b3a85cd] | 371 | self.dispersion = collections.OrderedDict() |
---|
[fa5fd8d] | 372 | self.details = {} |
---|
[8977226] | 373 | for p in self._model_info.parameters.user_parameters({}, is2d=True): |
---|
[04045f4] | 374 | if p.name in hidden: |
---|
[fa5fd8d] | 375 | continue |
---|
[fcd7bbd] | 376 | self.params[p.name] = p.default |
---|
[fa5fd8d] | 377 | self.details[p.id] = [p.units, p.limits[0], p.limits[1]] |
---|
[fb5914f] | 378 | if p.polydisperse: |
---|
[fa5fd8d] | 379 | self.details[p.id+".width"] = [ |
---|
| 380 | "", 0.0, 1.0 if p.relative_pd else np.inf |
---|
| 381 | ] |
---|
[fb5914f] | 382 | self.dispersion[p.name] = { |
---|
| 383 | 'width': 0, |
---|
| 384 | 'npts': 35, |
---|
| 385 | 'nsigmas': 3, |
---|
| 386 | 'type': 'gaussian', |
---|
| 387 | } |
---|
[ce27e21] | 388 | |
---|
[de97440] | 389 | def __get_state__(self): |
---|
[fa5fd8d] | 390 | # type: () -> Dict[str, Any] |
---|
[de97440] | 391 | state = self.__dict__.copy() |
---|
[4d76711] | 392 | state.pop('_model') |
---|
[de97440] | 393 | # May need to reload model info on set state since it has pointers |
---|
| 394 | # to python implementations of Iq, etc. |
---|
| 395 | #state.pop('_model_info') |
---|
| 396 | return state |
---|
| 397 | |
---|
| 398 | def __set_state__(self, state): |
---|
[fa5fd8d] | 399 | # type: (Dict[str, Any]) -> None |
---|
[de97440] | 400 | self.__dict__ = state |
---|
[fb5914f] | 401 | self._model = None |
---|
[de97440] | 402 | |
---|
[ce27e21] | 403 | def __str__(self): |
---|
[fa5fd8d] | 404 | # type: () -> str |
---|
[ce27e21] | 405 | """ |
---|
| 406 | :return: string representation |
---|
| 407 | """ |
---|
| 408 | return self.name |
---|
| 409 | |
---|
| 410 | def is_fittable(self, par_name): |
---|
[fa5fd8d] | 411 | # type: (str) -> bool |
---|
[ce27e21] | 412 | """ |
---|
| 413 | Check if a given parameter is fittable or not |
---|
| 414 | |
---|
| 415 | :param par_name: the parameter name to check |
---|
| 416 | """ |
---|
[e758662] | 417 | return par_name in self.fixed |
---|
[ce27e21] | 418 | #For the future |
---|
| 419 | #return self.params[str(par_name)].is_fittable() |
---|
| 420 | |
---|
| 421 | |
---|
| 422 | def getProfile(self): |
---|
[fa5fd8d] | 423 | # type: () -> (np.ndarray, np.ndarray) |
---|
[ce27e21] | 424 | """ |
---|
| 425 | Get SLD profile |
---|
| 426 | |
---|
| 427 | : return: (z, beta) where z is a list of depth of the transition points |
---|
| 428 | beta is a list of the corresponding SLD values |
---|
| 429 | """ |
---|
[745b7bb] | 430 | args = {} # type: Dict[str, Any] |
---|
[fa5fd8d] | 431 | for p in self._model_info.parameters.kernel_parameters: |
---|
| 432 | if p.id == self.multiplicity_info.control: |
---|
[745b7bb] | 433 | value = float(self.multiplicity) |
---|
[fa5fd8d] | 434 | elif p.length == 1: |
---|
[745b7bb] | 435 | value = self.params.get(p.id, np.NaN) |
---|
[fa5fd8d] | 436 | else: |
---|
[745b7bb] | 437 | value = np.array([self.params.get(p.id+str(k), np.NaN) |
---|
[b32dafd] | 438 | for k in range(1, p.length+1)]) |
---|
[745b7bb] | 439 | args[p.id] = value |
---|
| 440 | |
---|
[e7fe459] | 441 | x, y = self._model_info.profile(**args) |
---|
| 442 | return x, 1e-6*y |
---|
[ce27e21] | 443 | |
---|
| 444 | def setParam(self, name, value): |
---|
[fa5fd8d] | 445 | # type: (str, float) -> None |
---|
[ce27e21] | 446 | """ |
---|
| 447 | Set the value of a model parameter |
---|
| 448 | |
---|
| 449 | :param name: name of the parameter |
---|
| 450 | :param value: value of the parameter |
---|
| 451 | |
---|
| 452 | """ |
---|
| 453 | # Look for dispersion parameters |
---|
| 454 | toks = name.split('.') |
---|
[de0c4ba] | 455 | if len(toks) == 2: |
---|
[ce27e21] | 456 | for item in self.dispersion.keys(): |
---|
[e758662] | 457 | if item == toks[0]: |
---|
[ce27e21] | 458 | for par in self.dispersion[item]: |
---|
[e758662] | 459 | if par == toks[1]: |
---|
[ce27e21] | 460 | self.dispersion[item][par] = value |
---|
| 461 | return |
---|
| 462 | else: |
---|
| 463 | # Look for standard parameter |
---|
| 464 | for item in self.params.keys(): |
---|
[e758662] | 465 | if item == name: |
---|
[ce27e21] | 466 | self.params[item] = value |
---|
| 467 | return |
---|
| 468 | |
---|
[63b32bb] | 469 | raise ValueError("Model does not contain parameter %s" % name) |
---|
[ce27e21] | 470 | |
---|
| 471 | def getParam(self, name): |
---|
[fa5fd8d] | 472 | # type: (str) -> float |
---|
[ce27e21] | 473 | """ |
---|
| 474 | Set the value of a model parameter |
---|
| 475 | |
---|
| 476 | :param name: name of the parameter |
---|
| 477 | |
---|
| 478 | """ |
---|
| 479 | # Look for dispersion parameters |
---|
| 480 | toks = name.split('.') |
---|
[de0c4ba] | 481 | if len(toks) == 2: |
---|
[ce27e21] | 482 | for item in self.dispersion.keys(): |
---|
[e758662] | 483 | if item == toks[0]: |
---|
[ce27e21] | 484 | for par in self.dispersion[item]: |
---|
[e758662] | 485 | if par == toks[1]: |
---|
[ce27e21] | 486 | return self.dispersion[item][par] |
---|
| 487 | else: |
---|
| 488 | # Look for standard parameter |
---|
| 489 | for item in self.params.keys(): |
---|
[e758662] | 490 | if item == name: |
---|
[ce27e21] | 491 | return self.params[item] |
---|
| 492 | |
---|
[63b32bb] | 493 | raise ValueError("Model does not contain parameter %s" % name) |
---|
[ce27e21] | 494 | |
---|
| 495 | def getParamList(self): |
---|
[04dc697] | 496 | # type: () -> Sequence[str] |
---|
[ce27e21] | 497 | """ |
---|
| 498 | Return a list of all available parameters for the model |
---|
| 499 | """ |
---|
[04dc697] | 500 | param_list = list(self.params.keys()) |
---|
[ce27e21] | 501 | # WARNING: Extending the list with the dispersion parameters |
---|
[de0c4ba] | 502 | param_list.extend(self.getDispParamList()) |
---|
| 503 | return param_list |
---|
[ce27e21] | 504 | |
---|
| 505 | def getDispParamList(self): |
---|
[04dc697] | 506 | # type: () -> Sequence[str] |
---|
[ce27e21] | 507 | """ |
---|
[fb5914f] | 508 | Return a list of polydispersity parameters for the model |
---|
[ce27e21] | 509 | """ |
---|
[1780d59] | 510 | # TODO: fix test so that parameter order doesn't matter |
---|
[3bcb88c] | 511 | ret = ['%s.%s' % (p_name, ext) |
---|
| 512 | for p_name in self.dispersion.keys() |
---|
| 513 | for ext in ('npts', 'nsigmas', 'width')] |
---|
[9404dd3] | 514 | #print(ret) |
---|
[1780d59] | 515 | return ret |
---|
[ce27e21] | 516 | |
---|
| 517 | def clone(self): |
---|
[04dc697] | 518 | # type: () -> "SasviewModel" |
---|
[ce27e21] | 519 | """ Return a identical copy of self """ |
---|
| 520 | return deepcopy(self) |
---|
| 521 | |
---|
| 522 | def run(self, x=0.0): |
---|
[fa5fd8d] | 523 | # type: (Union[float, (float, float), List[float]]) -> float |
---|
[ce27e21] | 524 | """ |
---|
| 525 | Evaluate the model |
---|
| 526 | |
---|
| 527 | :param x: input q, or [q,phi] |
---|
| 528 | |
---|
| 529 | :return: scattering function P(q) |
---|
| 530 | |
---|
| 531 | **DEPRECATED**: use calculate_Iq instead |
---|
| 532 | """ |
---|
[de0c4ba] | 533 | if isinstance(x, (list, tuple)): |
---|
[3c56da87] | 534 | # pylint: disable=unpacking-non-sequence |
---|
[ce27e21] | 535 | q, phi = x |
---|
[60f03de] | 536 | return self.calculate_Iq([q*math.cos(phi)], [q*math.sin(phi)])[0] |
---|
[ce27e21] | 537 | else: |
---|
[60f03de] | 538 | return self.calculate_Iq([x])[0] |
---|
[ce27e21] | 539 | |
---|
| 540 | |
---|
| 541 | def runXY(self, x=0.0): |
---|
[fa5fd8d] | 542 | # type: (Union[float, (float, float), List[float]]) -> float |
---|
[ce27e21] | 543 | """ |
---|
| 544 | Evaluate the model in cartesian coordinates |
---|
| 545 | |
---|
| 546 | :param x: input q, or [qx, qy] |
---|
| 547 | |
---|
| 548 | :return: scattering function P(q) |
---|
| 549 | |
---|
| 550 | **DEPRECATED**: use calculate_Iq instead |
---|
| 551 | """ |
---|
[de0c4ba] | 552 | if isinstance(x, (list, tuple)): |
---|
[60f03de] | 553 | return self.calculate_Iq([x[0]], [x[1]])[0] |
---|
[ce27e21] | 554 | else: |
---|
[60f03de] | 555 | return self.calculate_Iq([x])[0] |
---|
[ce27e21] | 556 | |
---|
| 557 | def evalDistribution(self, qdist): |
---|
[04dc697] | 558 | # type: (Union[np.ndarray, Tuple[np.ndarray, np.ndarray], List[np.ndarray]]) -> np.ndarray |
---|
[d138d43] | 559 | r""" |
---|
[ce27e21] | 560 | Evaluate a distribution of q-values. |
---|
| 561 | |
---|
[d138d43] | 562 | :param qdist: array of q or a list of arrays [qx,qy] |
---|
[ce27e21] | 563 | |
---|
[d138d43] | 564 | * For 1D, a numpy array is expected as input |
---|
[ce27e21] | 565 | |
---|
[d138d43] | 566 | :: |
---|
[ce27e21] | 567 | |
---|
[d138d43] | 568 | evalDistribution(q) |
---|
[ce27e21] | 569 | |
---|
[d138d43] | 570 | where *q* is a numpy array. |
---|
[ce27e21] | 571 | |
---|
[d138d43] | 572 | * For 2D, a list of *[qx,qy]* is expected with 1D arrays as input |
---|
[ce27e21] | 573 | |
---|
[d138d43] | 574 | :: |
---|
[ce27e21] | 575 | |
---|
[d138d43] | 576 | qx = [ qx[0], qx[1], qx[2], ....] |
---|
| 577 | qy = [ qy[0], qy[1], qy[2], ....] |
---|
[ce27e21] | 578 | |
---|
[d138d43] | 579 | If the model is 1D only, then |
---|
[ce27e21] | 580 | |
---|
[d138d43] | 581 | .. math:: |
---|
[ce27e21] | 582 | |
---|
[d138d43] | 583 | q = \sqrt{q_x^2+q_y^2} |
---|
[ce27e21] | 584 | |
---|
| 585 | """ |
---|
[de0c4ba] | 586 | if isinstance(qdist, (list, tuple)): |
---|
[ce27e21] | 587 | # Check whether we have a list of ndarrays [qx,qy] |
---|
| 588 | qx, qy = qdist |
---|
[6d6508e] | 589 | if not self._model_info.parameters.has_2d: |
---|
[de0c4ba] | 590 | return self.calculate_Iq(np.sqrt(qx ** 2 + qy ** 2)) |
---|
[5d4777d] | 591 | else: |
---|
| 592 | return self.calculate_Iq(qx, qy) |
---|
[ce27e21] | 593 | |
---|
| 594 | elif isinstance(qdist, np.ndarray): |
---|
| 595 | # We have a simple 1D distribution of q-values |
---|
| 596 | return self.calculate_Iq(qdist) |
---|
| 597 | |
---|
| 598 | else: |
---|
[3c56da87] | 599 | raise TypeError("evalDistribution expects q or [qx, qy], not %r" |
---|
| 600 | % type(qdist)) |
---|
[ce27e21] | 601 | |
---|
[9dcb21d] | 602 | def calc_composition_models(self, qx): |
---|
[64614ad] | 603 | """ |
---|
[9dcb21d] | 604 | returns parts of the composition model or None if not a composition |
---|
| 605 | model. |
---|
[64614ad] | 606 | """ |
---|
[946c8d27] | 607 | # TODO: have calculate_Iq return the intermediates. |
---|
| 608 | # |
---|
| 609 | # The current interface causes calculate_Iq() to be called twice, |
---|
| 610 | # once to get the combined result and again to get the intermediate |
---|
| 611 | # results. This is necessary for now. |
---|
| 612 | # Long term, the solution is to change the interface to calculate_Iq |
---|
| 613 | # so that it returns a results object containing all the bits: |
---|
[9644b5a] | 614 | # the A, B, C, ... of the composition model (and any subcomponents?) |
---|
[946c8d27] | 615 | # the P and S of the product model, |
---|
| 616 | # the combined model before resolution smearing, |
---|
| 617 | # the sasmodel before sesans conversion, |
---|
| 618 | # the oriented 2D model used to fit oriented usans data, |
---|
| 619 | # the final I(q), |
---|
| 620 | # ... |
---|
[9644b5a] | 621 | # |
---|
[946c8d27] | 622 | # Have the model calculator add all of these blindly to the data |
---|
| 623 | # tree, and update the graphs which contain them. The fitter |
---|
| 624 | # needs to be updated to use the I(q) value only, ignoring the rest. |
---|
| 625 | # |
---|
| 626 | # The simple fix of returning the existing intermediate results |
---|
| 627 | # will not work for a couple of reasons: (1) another thread may |
---|
| 628 | # sneak in to compute its own results before calc_composition_models |
---|
| 629 | # is called, and (2) calculate_Iq is currently called three times: |
---|
| 630 | # once with q, once with q values before qmin and once with q values |
---|
| 631 | # after q max. Both of these should be addressed before |
---|
| 632 | # replacing this code. |
---|
[9644b5a] | 633 | composition = self._model_info.composition |
---|
| 634 | if composition and composition[0] == 'product': # only P*S for now |
---|
| 635 | with calculation_lock: |
---|
| 636 | self._calculate_Iq(qx) |
---|
| 637 | return self._intermediate_results |
---|
| 638 | else: |
---|
| 639 | return None |
---|
[bf8c271] | 640 | |
---|
[fa5fd8d] | 641 | def calculate_Iq(self, qx, qy=None): |
---|
| 642 | # type: (Sequence[float], Optional[Sequence[float]]) -> np.ndarray |
---|
[ff7119b] | 643 | """ |
---|
| 644 | Calculate Iq for one set of q with the current parameters. |
---|
| 645 | |
---|
| 646 | If the model is 1D, use *q*. If 2D, use *qx*, *qy*. |
---|
| 647 | |
---|
| 648 | This should NOT be used for fitting since it copies the *q* vectors |
---|
| 649 | to the card for each evaluation. |
---|
| 650 | """ |
---|
[a38b065] | 651 | ## uncomment the following when trying to debug the uncoordinated calls |
---|
| 652 | ## to calculate_Iq |
---|
| 653 | #if calculation_lock.locked(): |
---|
[724257c] | 654 | # logger.info("calculation waiting for another thread to complete") |
---|
| 655 | # logger.info("\n".join(traceback.format_stack())) |
---|
[a38b065] | 656 | |
---|
| 657 | with calculation_lock: |
---|
| 658 | return self._calculate_Iq(qx, qy) |
---|
| 659 | |
---|
| 660 | def _calculate_Iq(self, qx, qy=None): |
---|
[6a0d6aa] | 661 | #core.HAVE_OPENCL = False |
---|
[fb5914f] | 662 | if self._model is None: |
---|
[d2bb604] | 663 | self._model = core.build_model(self._model_info) |
---|
[fa5fd8d] | 664 | if qy is not None: |
---|
| 665 | q_vectors = [np.asarray(qx), np.asarray(qy)] |
---|
| 666 | else: |
---|
| 667 | q_vectors = [np.asarray(qx)] |
---|
[a738209] | 668 | calculator = self._model.make_kernel(q_vectors) |
---|
[6a0d6aa] | 669 | parameters = self._model_info.parameters |
---|
| 670 | pairs = [self._get_weights(p) for p in parameters.call_parameters] |
---|
[9c1a59c] | 671 | #weights.plot_weights(self._model_info, pairs) |
---|
[bde38b5] | 672 | call_details, values, is_magnetic = make_kernel_args(calculator, pairs) |
---|
[4edec6f] | 673 | #call_details.show() |
---|
| 674 | #print("pairs", pairs) |
---|
| 675 | #print("params", self.params) |
---|
| 676 | #print("values", values) |
---|
| 677 | #print("is_mag", is_magnetic) |
---|
[6a0d6aa] | 678 | result = calculator(call_details, values, cutoff=self.cutoff, |
---|
[9eb3632] | 679 | magnetic=is_magnetic) |
---|
[bf8c271] | 680 | self._intermediate_results = getattr(calculator, 'results', None) |
---|
[a738209] | 681 | calculator.release() |
---|
[9f37726] | 682 | self._model.release() |
---|
[ce27e21] | 683 | return result |
---|
| 684 | |
---|
| 685 | def calculate_ER(self): |
---|
[fa5fd8d] | 686 | # type: () -> float |
---|
[ce27e21] | 687 | """ |
---|
| 688 | Calculate the effective radius for P(q)*S(q) |
---|
| 689 | |
---|
| 690 | :return: the value of the effective radius |
---|
| 691 | """ |
---|
[4bfd277] | 692 | if self._model_info.ER is None: |
---|
[ce27e21] | 693 | return 1.0 |
---|
| 694 | else: |
---|
[4bfd277] | 695 | value, weight = self._dispersion_mesh() |
---|
| 696 | fv = self._model_info.ER(*value) |
---|
[9404dd3] | 697 | #print(values[0].shape, weights.shape, fv.shape) |
---|
[4bfd277] | 698 | return np.sum(weight * fv) / np.sum(weight) |
---|
[ce27e21] | 699 | |
---|
| 700 | def calculate_VR(self): |
---|
[fa5fd8d] | 701 | # type: () -> float |
---|
[ce27e21] | 702 | """ |
---|
| 703 | Calculate the volf ratio for P(q)*S(q) |
---|
| 704 | |
---|
| 705 | :return: the value of the volf ratio |
---|
| 706 | """ |
---|
[4bfd277] | 707 | if self._model_info.VR is None: |
---|
[ce27e21] | 708 | return 1.0 |
---|
| 709 | else: |
---|
[4bfd277] | 710 | value, weight = self._dispersion_mesh() |
---|
| 711 | whole, part = self._model_info.VR(*value) |
---|
| 712 | return np.sum(weight * part) / np.sum(weight * whole) |
---|
[ce27e21] | 713 | |
---|
| 714 | def set_dispersion(self, parameter, dispersion): |
---|
[fa5fd8d] | 715 | # type: (str, weights.Dispersion) -> Dict[str, Any] |
---|
[ce27e21] | 716 | """ |
---|
| 717 | Set the dispersion object for a model parameter |
---|
| 718 | |
---|
| 719 | :param parameter: name of the parameter [string] |
---|
| 720 | :param dispersion: dispersion object of type Dispersion |
---|
| 721 | """ |
---|
[fa800e72] | 722 | if parameter in self.params: |
---|
[1780d59] | 723 | # TODO: Store the disperser object directly in the model. |
---|
[56b2687] | 724 | # The current method of relying on the sasview GUI to |
---|
[fa800e72] | 725 | # remember them is kind of funky. |
---|
[1780d59] | 726 | # Note: can't seem to get disperser parameters from sasview |
---|
[9c1a59c] | 727 | # (1) Could create a sasview model that has not yet been |
---|
[1780d59] | 728 | # converted, assign the disperser to one of its polydisperse |
---|
| 729 | # parameters, then retrieve the disperser parameters from the |
---|
[9c1a59c] | 730 | # sasview model. |
---|
| 731 | # (2) Could write a disperser parameter retriever in sasview. |
---|
| 732 | # (3) Could modify sasview to use sasmodels.weights dispersers. |
---|
[1780d59] | 733 | # For now, rely on the fact that the sasview only ever uses |
---|
| 734 | # new dispersers in the set_dispersion call and create a new |
---|
| 735 | # one instead of trying to assign parameters. |
---|
[ce27e21] | 736 | self.dispersion[parameter] = dispersion.get_pars() |
---|
| 737 | else: |
---|
| 738 | raise ValueError("%r is not a dispersity or orientation parameter") |
---|
| 739 | |
---|
[aa4946b] | 740 | def _dispersion_mesh(self): |
---|
[fa5fd8d] | 741 | # type: () -> List[Tuple[np.ndarray, np.ndarray]] |
---|
[ce27e21] | 742 | """ |
---|
| 743 | Create a mesh grid of dispersion parameters and weights. |
---|
| 744 | |
---|
| 745 | Returns [p1,p2,...],w where pj is a vector of values for parameter j |
---|
| 746 | and w is a vector containing the products for weights for each |
---|
| 747 | parameter set in the vector. |
---|
| 748 | """ |
---|
[4bfd277] | 749 | pars = [self._get_weights(p) |
---|
| 750 | for p in self._model_info.parameters.call_parameters |
---|
| 751 | if p.type == 'volume'] |
---|
[9eb3632] | 752 | return dispersion_mesh(self._model_info, pars) |
---|
[ce27e21] | 753 | |
---|
| 754 | def _get_weights(self, par): |
---|
[fa5fd8d] | 755 | # type: (Parameter) -> Tuple[np.ndarray, np.ndarray] |
---|
[de0c4ba] | 756 | """ |
---|
[fb5914f] | 757 | Return dispersion weights for parameter |
---|
[de0c4ba] | 758 | """ |
---|
[fa5fd8d] | 759 | if par.name not in self.params: |
---|
| 760 | if par.name == self.multiplicity_info.control: |
---|
[4edec6f] | 761 | return [self.multiplicity], [1.0] |
---|
[fa5fd8d] | 762 | else: |
---|
[8f93522] | 763 | # For hidden parameters use the default value. |
---|
| 764 | value = self._model_info.parameters.defaults.get(par.name, np.NaN) |
---|
| 765 | return [value], [1.0] |
---|
[fa5fd8d] | 766 | elif par.polydisperse: |
---|
[fb5914f] | 767 | dis = self.dispersion[par.name] |
---|
[9c1a59c] | 768 | if dis['type'] == 'array': |
---|
| 769 | value, weight = dis['values'], dis['weights'] |
---|
| 770 | else: |
---|
| 771 | value, weight = weights.get_weights( |
---|
| 772 | dis['type'], dis['npts'], dis['width'], dis['nsigmas'], |
---|
| 773 | self.params[par.name], par.limits, par.relative_pd) |
---|
[fb5914f] | 774 | return value, weight / np.sum(weight) |
---|
| 775 | else: |
---|
[4edec6f] | 776 | return [self.params[par.name]], [1.0] |
---|
[ce27e21] | 777 | |
---|
[749a7d4] | 778 | def test_cylinder(): |
---|
[fa5fd8d] | 779 | # type: () -> float |
---|
[4d76711] | 780 | """ |
---|
[749a7d4] | 781 | Test that the cylinder model runs, returning the value at [0.1,0.1]. |
---|
[4d76711] | 782 | """ |
---|
| 783 | Cylinder = _make_standard_model('cylinder') |
---|
[fb5914f] | 784 | cylinder = Cylinder() |
---|
[b32dafd] | 785 | return cylinder.evalDistribution([0.1, 0.1]) |
---|
[de97440] | 786 | |
---|
[8f93522] | 787 | def test_structure_factor(): |
---|
| 788 | # type: () -> float |
---|
| 789 | """ |
---|
[749a7d4] | 790 | Test that 2-D hardsphere model runs and doesn't produce NaN. |
---|
[8f93522] | 791 | """ |
---|
| 792 | Model = _make_standard_model('hardsphere') |
---|
| 793 | model = Model() |
---|
| 794 | value = model.evalDistribution([0.1, 0.1]) |
---|
| 795 | if np.isnan(value): |
---|
| 796 | raise ValueError("hardsphere returns null") |
---|
| 797 | |
---|
[04045f4] | 798 | def test_rpa(): |
---|
| 799 | # type: () -> float |
---|
| 800 | """ |
---|
[749a7d4] | 801 | Test that the 2-D RPA model runs |
---|
[04045f4] | 802 | """ |
---|
| 803 | RPA = _make_standard_model('rpa') |
---|
| 804 | rpa = RPA(3) |
---|
[b32dafd] | 805 | return rpa.evalDistribution([0.1, 0.1]) |
---|
[04045f4] | 806 | |
---|
[749a7d4] | 807 | def test_empty_distribution(): |
---|
| 808 | # type: () -> None |
---|
| 809 | """ |
---|
| 810 | Make sure that sasmodels returns NaN when there are no polydispersity points |
---|
| 811 | """ |
---|
| 812 | Cylinder = _make_standard_model('cylinder') |
---|
| 813 | cylinder = Cylinder() |
---|
| 814 | cylinder.setParam('radius', -1.0) |
---|
| 815 | cylinder.setParam('background', 0.) |
---|
| 816 | Iq = cylinder.evalDistribution(np.asarray([0.1])) |
---|
| 817 | assert np.isnan(Iq[0]), "empty distribution fails" |
---|
[4d76711] | 818 | |
---|
| 819 | def test_model_list(): |
---|
[fa5fd8d] | 820 | # type: () -> None |
---|
[4d76711] | 821 | """ |
---|
[749a7d4] | 822 | Make sure that all models build as sasview models |
---|
[4d76711] | 823 | """ |
---|
| 824 | from .exception import annotate_exception |
---|
| 825 | for name in core.list_models(): |
---|
| 826 | try: |
---|
| 827 | _make_standard_model(name) |
---|
| 828 | except: |
---|
| 829 | annotate_exception("when loading "+name) |
---|
| 830 | raise |
---|
| 831 | |
---|
[c95dfc63] | 832 | def test_old_name(): |
---|
| 833 | # type: () -> None |
---|
| 834 | """ |
---|
| 835 | Load and run cylinder model from sas.models.CylinderModel |
---|
| 836 | """ |
---|
| 837 | if not SUPPORT_OLD_STYLE_PLUGINS: |
---|
| 838 | return |
---|
| 839 | try: |
---|
| 840 | # if sasview is not on the path then don't try to test it |
---|
| 841 | import sas |
---|
| 842 | except ImportError: |
---|
| 843 | return |
---|
| 844 | load_standard_models() |
---|
| 845 | from sas.models.CylinderModel import CylinderModel |
---|
| 846 | CylinderModel().evalDistribution([0.1, 0.1]) |
---|
| 847 | |
---|
[fb5914f] | 848 | if __name__ == "__main__": |
---|
[749a7d4] | 849 | print("cylinder(0.1,0.1)=%g"%test_cylinder()) |
---|
| 850 | #test_empty_distribution() |
---|