[51f14603] | 1 | |
---|
| 2 | import copy |
---|
| 3 | #import logging |
---|
| 4 | import sys |
---|
| 5 | import math |
---|
[6fe5100] | 6 | import numpy |
---|
| 7 | |
---|
[79492222] | 8 | from sas.dataloader.data_info import Data1D |
---|
| 9 | from sas.dataloader.data_info import Data2D |
---|
[6fe5100] | 10 | _SMALLVALUE = 1.0e-10 |
---|
| 11 | |
---|
| 12 | class FitHandler(object): |
---|
[51f14603] | 13 | """ |
---|
[6fe5100] | 14 | Abstract interface for fit thread handler. |
---|
| 15 | |
---|
| 16 | The methods in this class are called by the optimizer as the fit |
---|
| 17 | progresses. |
---|
| 18 | |
---|
| 19 | Note that it is up to the optimizer to call the fit handler correctly, |
---|
| 20 | reporting all status changes and maintaining the 'done' flag. |
---|
[51f14603] | 21 | """ |
---|
[6fe5100] | 22 | done = False |
---|
| 23 | """True when the fit job is complete""" |
---|
| 24 | result = None |
---|
| 25 | """The current best result of the fit""" |
---|
| 26 | |
---|
| 27 | def improvement(self): |
---|
[51f14603] | 28 | """ |
---|
[6fe5100] | 29 | Called when a result is observed which is better than previous |
---|
| 30 | results from the fit. |
---|
| 31 | |
---|
| 32 | result is a FitResult object, with parameters, #calls and fitness. |
---|
[51f14603] | 33 | """ |
---|
[6fe5100] | 34 | def error(self, msg): |
---|
[51f14603] | 35 | """ |
---|
[6fe5100] | 36 | Model had an error; print traceback |
---|
[51f14603] | 37 | """ |
---|
[6fe5100] | 38 | def progress(self, current, expected): |
---|
[51f14603] | 39 | """ |
---|
[6fe5100] | 40 | Called each cycle of the fit, reporting the current and the |
---|
| 41 | expected amount of work. The meaning of these values is |
---|
| 42 | optimizer dependent, but they can be converted into a percent |
---|
| 43 | complete using (100*current)//expected. |
---|
| 44 | |
---|
| 45 | Progress is updated each iteration of the fit, whatever that |
---|
| 46 | means for the particular optimization algorithm. It is called |
---|
| 47 | after any calls to improvement for the iteration so that the |
---|
| 48 | update handler can control I/O bandwidth by suppressing |
---|
| 49 | intermediate improvements until the fit is complete. |
---|
[51f14603] | 50 | """ |
---|
[6fe5100] | 51 | def finalize(self): |
---|
[51f14603] | 52 | """ |
---|
[6fe5100] | 53 | Fit is complete; best results are reported |
---|
[51f14603] | 54 | """ |
---|
[6fe5100] | 55 | def abort(self): |
---|
[51f14603] | 56 | """ |
---|
[6fe5100] | 57 | Fit was aborted. |
---|
[51f14603] | 58 | """ |
---|
[6fe5100] | 59 | |
---|
[95d58d3] | 60 | # TODO: not sure how these are used, but they are needed for running the fit |
---|
| 61 | def update_fit(self, last=False): pass |
---|
| 62 | def set_result(self, result=None): self.result = result |
---|
| 63 | |
---|
[6fe5100] | 64 | class Model: |
---|
[51f14603] | 65 | """ |
---|
[fd5ac0d] | 66 | Fit wrapper for SAS models. |
---|
[51f14603] | 67 | """ |
---|
[fd5ac0d] | 68 | def __init__(self, sas_model, sas_data=None, **kw): |
---|
[51f14603] | 69 | """ |
---|
[386ffe1] | 70 | :param sas_model: the sas model to wrap for fitting |
---|
[6fe5100] | 71 | |
---|
[51f14603] | 72 | """ |
---|
[fd5ac0d] | 73 | self.model = sas_model |
---|
| 74 | self.name = sas_model.name |
---|
| 75 | self.data = sas_data |
---|
[6fe5100] | 76 | |
---|
[51f14603] | 77 | def get_params(self, fitparams): |
---|
| 78 | """ |
---|
| 79 | return a list of value of paramter to fit |
---|
[6fe5100] | 80 | |
---|
[51f14603] | 81 | :param fitparams: list of paramaters name to fit |
---|
[6fe5100] | 82 | |
---|
[51f14603] | 83 | """ |
---|
[6fe5100] | 84 | return [self.model.getParam(k) for k in fitparams] |
---|
| 85 | |
---|
[51f14603] | 86 | def set_params(self, paramlist, params): |
---|
| 87 | """ |
---|
| 88 | Set value for parameters to fit |
---|
[6fe5100] | 89 | |
---|
[51f14603] | 90 | :param params: list of value for parameters to fit |
---|
[6fe5100] | 91 | |
---|
[51f14603] | 92 | """ |
---|
[6fe5100] | 93 | for k,v in zip(paramlist, params): |
---|
| 94 | self.model.setParam(k,v) |
---|
| 95 | |
---|
| 96 | def set(self, **kw): |
---|
| 97 | self.set_params(*zip(*kw.items())) |
---|
| 98 | |
---|
[51f14603] | 99 | def eval(self, x): |
---|
| 100 | """ |
---|
[386ffe1] | 101 | Override eval method of model. |
---|
[6fe5100] | 102 | |
---|
[51f14603] | 103 | :param x: the x value used to compute a function |
---|
| 104 | """ |
---|
| 105 | try: |
---|
| 106 | return self.model.evalDistribution(x) |
---|
| 107 | except: |
---|
| 108 | raise |
---|
[6fe5100] | 109 | |
---|
[51f14603] | 110 | def eval_derivs(self, x, pars=[]): |
---|
| 111 | """ |
---|
| 112 | Evaluate the model and derivatives wrt pars at x. |
---|
| 113 | |
---|
| 114 | pars is a list of the names of the parameters for which derivatives |
---|
| 115 | are desired. |
---|
| 116 | |
---|
| 117 | This method needs to be specialized in the model to evaluate the |
---|
| 118 | model function. Alternatively, the model can implement is own |
---|
| 119 | version of residuals which calculates the residuals directly |
---|
| 120 | instead of calling eval. |
---|
| 121 | """ |
---|
[6fe5100] | 122 | raise NotImplementedError('no derivatives available') |
---|
| 123 | |
---|
| 124 | def __call__(self, x): |
---|
| 125 | return self.eval(x) |
---|
[51f14603] | 126 | |
---|
| 127 | class FitData1D(Data1D): |
---|
| 128 | """ |
---|
[fd5ac0d] | 129 | Wrapper class for SAS data |
---|
[51f14603] | 130 | FitData1D inherits from DataLoader.data_info.Data1D. Implements |
---|
| 131 | a way to get residuals from data. |
---|
| 132 | """ |
---|
| 133 | def __init__(self, x, y, dx=None, dy=None, smearer=None, data=None): |
---|
| 134 | """ |
---|
| 135 | :param smearer: is an object of class QSmearer or SlitSmearer |
---|
| 136 | that will smear the theory data (slit smearing or resolution |
---|
| 137 | smearing) when set. |
---|
| 138 | |
---|
| 139 | The proper way to set the smearing object would be to |
---|
| 140 | do the following: :: |
---|
| 141 | |
---|
[79492222] | 142 | from sas.models.qsmearing import smear_selection |
---|
[51f14603] | 143 | smearer = smear_selection(some_data) |
---|
| 144 | fitdata1d = FitData1D( x= [1,3,..,], |
---|
| 145 | y= [3,4,..,8], |
---|
| 146 | dx=None, |
---|
| 147 | dy=[1,2...], smearer= smearer) |
---|
| 148 | |
---|
| 149 | :Note: that some_data _HAS_ to be of |
---|
| 150 | class DataLoader.data_info.Data1D |
---|
| 151 | Setting it back to None will turn smearing off. |
---|
| 152 | |
---|
| 153 | """ |
---|
| 154 | Data1D.__init__(self, x=x, y=y, dx=dx, dy=dy) |
---|
[6fe5100] | 155 | self.num_points = len(x) |
---|
[fd5ac0d] | 156 | self.sas_data = data |
---|
[51f14603] | 157 | self.smearer = smearer |
---|
| 158 | self._first_unsmeared_bin = None |
---|
| 159 | self._last_unsmeared_bin = None |
---|
| 160 | # Check error bar; if no error bar found, set it constant(=1) |
---|
| 161 | # TODO: Should provide an option for users to set it like percent, |
---|
| 162 | # constant, or dy data |
---|
[9f7fbd9] | 163 | if dy is None or dy == [] or dy.all() == 0: |
---|
[51f14603] | 164 | self.dy = numpy.ones(len(y)) |
---|
| 165 | else: |
---|
| 166 | self.dy = numpy.asarray(dy).copy() |
---|
| 167 | |
---|
| 168 | ## Min Q-value |
---|
| 169 | #Skip the Q=0 point, especially when y(q=0)=None at x[0]. |
---|
| 170 | if min(self.x) == 0.0 and self.x[0] == 0 and\ |
---|
| 171 | not numpy.isfinite(self.y[0]): |
---|
| 172 | self.qmin = min(self.x[self.x != 0]) |
---|
| 173 | else: |
---|
| 174 | self.qmin = min(self.x) |
---|
| 175 | ## Max Q-value |
---|
| 176 | self.qmax = max(self.x) |
---|
| 177 | |
---|
| 178 | # Range used for input to smearing |
---|
| 179 | self._qmin_unsmeared = self.qmin |
---|
| 180 | self._qmax_unsmeared = self.qmax |
---|
| 181 | # Identify the bin range for the unsmeared and smeared spaces |
---|
| 182 | self.idx = (self.x >= self.qmin) & (self.x <= self.qmax) |
---|
| 183 | self.idx_unsmeared = (self.x >= self._qmin_unsmeared) \ |
---|
| 184 | & (self.x <= self._qmax_unsmeared) |
---|
| 185 | |
---|
| 186 | def set_fit_range(self, qmin=None, qmax=None): |
---|
| 187 | """ to set the fit range""" |
---|
| 188 | # Skip Q=0 point, (especially for y(q=0)=None at x[0]). |
---|
| 189 | # ToDo: Find better way to do it. |
---|
| 190 | if qmin == 0.0 and not numpy.isfinite(self.y[qmin]): |
---|
| 191 | self.qmin = min(self.x[self.x != 0]) |
---|
| 192 | elif qmin != None: |
---|
| 193 | self.qmin = qmin |
---|
| 194 | if qmax != None: |
---|
| 195 | self.qmax = qmax |
---|
| 196 | # Determine the range needed in unsmeared-Q to cover |
---|
| 197 | # the smeared Q range |
---|
| 198 | self._qmin_unsmeared = self.qmin |
---|
| 199 | self._qmax_unsmeared = self.qmax |
---|
| 200 | |
---|
| 201 | self._first_unsmeared_bin = 0 |
---|
| 202 | self._last_unsmeared_bin = len(self.x) - 1 |
---|
| 203 | |
---|
| 204 | if self.smearer != None: |
---|
| 205 | self._first_unsmeared_bin, self._last_unsmeared_bin = \ |
---|
| 206 | self.smearer.get_bin_range(self.qmin, self.qmax) |
---|
| 207 | self._qmin_unsmeared = self.x[self._first_unsmeared_bin] |
---|
| 208 | self._qmax_unsmeared = self.x[self._last_unsmeared_bin] |
---|
| 209 | |
---|
| 210 | # Identify the bin range for the unsmeared and smeared spaces |
---|
| 211 | self.idx = (self.x >= self.qmin) & (self.x <= self.qmax) |
---|
| 212 | ## zero error can not participate for fitting |
---|
| 213 | self.idx = self.idx & (self.dy != 0) |
---|
| 214 | self.idx_unsmeared = (self.x >= self._qmin_unsmeared) \ |
---|
| 215 | & (self.x <= self._qmax_unsmeared) |
---|
| 216 | |
---|
| 217 | def get_fit_range(self): |
---|
| 218 | """ |
---|
| 219 | Return the range of data.x to fit |
---|
| 220 | """ |
---|
| 221 | return self.qmin, self.qmax |
---|
[95d58d3] | 222 | |
---|
| 223 | def size(self): |
---|
| 224 | """ |
---|
| 225 | Number of measurement points in data set after masking, etc. |
---|
| 226 | """ |
---|
| 227 | return len(self.x) |
---|
| 228 | |
---|
[51f14603] | 229 | def residuals(self, fn): |
---|
| 230 | """ |
---|
| 231 | Compute residuals. |
---|
| 232 | |
---|
| 233 | If self.smearer has been set, use if to smear |
---|
| 234 | the data before computing chi squared. |
---|
| 235 | |
---|
| 236 | :param fn: function that return model value |
---|
| 237 | |
---|
| 238 | :return: residuals |
---|
| 239 | """ |
---|
| 240 | # Compute theory data f(x) |
---|
| 241 | fx = numpy.zeros(len(self.x)) |
---|
| 242 | fx[self.idx_unsmeared] = fn(self.x[self.idx_unsmeared]) |
---|
| 243 | |
---|
| 244 | ## Smear theory data |
---|
| 245 | if self.smearer is not None: |
---|
| 246 | fx = self.smearer(fx, self._first_unsmeared_bin, |
---|
| 247 | self._last_unsmeared_bin) |
---|
| 248 | ## Sanity check |
---|
| 249 | if numpy.size(self.dy) != numpy.size(fx): |
---|
| 250 | msg = "FitData1D: invalid error array " |
---|
| 251 | msg += "%d <> %d" % (numpy.shape(self.dy), numpy.size(fx)) |
---|
| 252 | raise RuntimeError, msg |
---|
| 253 | return (self.y[self.idx] - fx[self.idx]) / self.dy[self.idx], fx[self.idx] |
---|
| 254 | |
---|
| 255 | def residuals_deriv(self, model, pars=[]): |
---|
| 256 | """ |
---|
| 257 | :return: residuals derivatives . |
---|
| 258 | |
---|
| 259 | :note: in this case just return empty array |
---|
| 260 | """ |
---|
| 261 | return [] |
---|
[9f7fbd9] | 262 | |
---|
| 263 | |
---|
[51f14603] | 264 | class FitData2D(Data2D): |
---|
| 265 | """ |
---|
[fd5ac0d] | 266 | Wrapper class for SAS data |
---|
[51f14603] | 267 | """ |
---|
[fd5ac0d] | 268 | def __init__(self, sas_data2d, data=None, err_data=None): |
---|
[51f14603] | 269 | Data2D.__init__(self, data=data, err_data=err_data) |
---|
[79492222] | 270 | # Data can be initialized with a sas plottable or with vectors. |
---|
[51f14603] | 271 | self.res_err_image = [] |
---|
[95d58d3] | 272 | self.num_points = 0 # will be set by set_data |
---|
[51f14603] | 273 | self.idx = [] |
---|
| 274 | self.qmin = None |
---|
| 275 | self.qmax = None |
---|
| 276 | self.smearer = None |
---|
| 277 | self.radius = 0 |
---|
| 278 | self.res_err_data = [] |
---|
[fd5ac0d] | 279 | self.sas_data = sas_data2d |
---|
| 280 | self.set_data(sas_data2d) |
---|
[51f14603] | 281 | |
---|
[fd5ac0d] | 282 | def set_data(self, sas_data2d, qmin=None, qmax=None): |
---|
[51f14603] | 283 | """ |
---|
| 284 | Determine the correct qx_data and qy_data within range to fit |
---|
| 285 | """ |
---|
[fd5ac0d] | 286 | self.data = sas_data2d.data |
---|
| 287 | self.err_data = sas_data2d.err_data |
---|
| 288 | self.qx_data = sas_data2d.qx_data |
---|
| 289 | self.qy_data = sas_data2d.qy_data |
---|
| 290 | self.mask = sas_data2d.mask |
---|
[51f14603] | 291 | |
---|
[fd5ac0d] | 292 | x_max = max(math.fabs(sas_data2d.xmin), math.fabs(sas_data2d.xmax)) |
---|
| 293 | y_max = max(math.fabs(sas_data2d.ymin), math.fabs(sas_data2d.ymax)) |
---|
[51f14603] | 294 | |
---|
| 295 | ## fitting range |
---|
| 296 | if qmin == None: |
---|
| 297 | self.qmin = 1e-16 |
---|
| 298 | if qmax == None: |
---|
| 299 | self.qmax = math.sqrt(x_max * x_max + y_max * y_max) |
---|
| 300 | ## new error image for fitting purpose |
---|
| 301 | if self.err_data == None or self.err_data == []: |
---|
| 302 | self.res_err_data = numpy.ones(len(self.data)) |
---|
| 303 | else: |
---|
| 304 | self.res_err_data = copy.deepcopy(self.err_data) |
---|
| 305 | #self.res_err_data[self.res_err_data==0]=1 |
---|
| 306 | |
---|
| 307 | self.radius = numpy.sqrt(self.qx_data**2 + self.qy_data**2) |
---|
| 308 | |
---|
| 309 | # Note: mask = True: for MASK while mask = False for NOT to mask |
---|
| 310 | self.idx = ((self.qmin <= self.radius) &\ |
---|
| 311 | (self.radius <= self.qmax)) |
---|
| 312 | self.idx = (self.idx) & (self.mask) |
---|
| 313 | self.idx = (self.idx) & (numpy.isfinite(self.data)) |
---|
[95d58d3] | 314 | self.num_points = numpy.sum(self.idx) |
---|
[51f14603] | 315 | |
---|
| 316 | def set_smearer(self, smearer): |
---|
| 317 | """ |
---|
| 318 | Set smearer |
---|
| 319 | """ |
---|
| 320 | if smearer == None: |
---|
| 321 | return |
---|
| 322 | self.smearer = smearer |
---|
| 323 | self.smearer.set_index(self.idx) |
---|
| 324 | self.smearer.get_data() |
---|
| 325 | |
---|
| 326 | def set_fit_range(self, qmin=None, qmax=None): |
---|
| 327 | """ |
---|
| 328 | To set the fit range |
---|
| 329 | """ |
---|
| 330 | if qmin == 0.0: |
---|
| 331 | self.qmin = 1e-16 |
---|
| 332 | elif qmin != None: |
---|
| 333 | self.qmin = qmin |
---|
| 334 | if qmax != None: |
---|
| 335 | self.qmax = qmax |
---|
| 336 | self.radius = numpy.sqrt(self.qx_data**2 + self.qy_data**2) |
---|
| 337 | self.idx = ((self.qmin <= self.radius) &\ |
---|
| 338 | (self.radius <= self.qmax)) |
---|
| 339 | self.idx = (self.idx) & (self.mask) |
---|
| 340 | self.idx = (self.idx) & (numpy.isfinite(self.data)) |
---|
| 341 | self.idx = (self.idx) & (self.res_err_data != 0) |
---|
| 342 | |
---|
| 343 | def get_fit_range(self): |
---|
| 344 | """ |
---|
| 345 | return the range of data.x to fit |
---|
| 346 | """ |
---|
| 347 | return self.qmin, self.qmax |
---|
[95d58d3] | 348 | |
---|
| 349 | def size(self): |
---|
| 350 | """ |
---|
| 351 | Number of measurement points in data set after masking, etc. |
---|
| 352 | """ |
---|
| 353 | return numpy.sum(self.idx) |
---|
| 354 | |
---|
[51f14603] | 355 | def residuals(self, fn): |
---|
| 356 | """ |
---|
| 357 | return the residuals |
---|
| 358 | """ |
---|
| 359 | if self.smearer != None: |
---|
| 360 | fn.set_index(self.idx) |
---|
| 361 | # Get necessary data from self.data and set the data for smearing |
---|
| 362 | fn.get_data() |
---|
| 363 | |
---|
| 364 | gn = fn.get_value() |
---|
| 365 | else: |
---|
| 366 | gn = fn([self.qx_data[self.idx], |
---|
| 367 | self.qy_data[self.idx]]) |
---|
| 368 | # use only the data point within ROI range |
---|
| 369 | res = (self.data[self.idx] - gn) / self.res_err_data[self.idx] |
---|
| 370 | |
---|
| 371 | return res, gn |
---|
| 372 | |
---|
| 373 | def residuals_deriv(self, model, pars=[]): |
---|
| 374 | """ |
---|
| 375 | :return: residuals derivatives . |
---|
| 376 | |
---|
| 377 | :note: in this case just return empty array |
---|
| 378 | |
---|
| 379 | """ |
---|
| 380 | return [] |
---|
| 381 | |
---|
| 382 | |
---|
| 383 | class FitAbort(Exception): |
---|
| 384 | """ |
---|
| 385 | Exception raise to stop the fit |
---|
| 386 | """ |
---|
| 387 | #pass |
---|
| 388 | #print"Creating fit abort Exception" |
---|
| 389 | |
---|
| 390 | |
---|
| 391 | |
---|
| 392 | class FitEngine: |
---|
| 393 | def __init__(self): |
---|
| 394 | """ |
---|
[386ffe1] | 395 | Base class for the fit engine |
---|
[51f14603] | 396 | """ |
---|
| 397 | #Dictionnary of fitArrange element (fit problems) |
---|
| 398 | self.fit_arrange_dict = {} |
---|
| 399 | self.fitter_id = None |
---|
| 400 | |
---|
| 401 | def set_model(self, model, id, pars=[], constraints=[], data=None): |
---|
| 402 | """ |
---|
| 403 | set a model on a given in the fit engine. |
---|
| 404 | |
---|
[79492222] | 405 | :param model: sas.models type |
---|
[51f14603] | 406 | :param id: is the key of the fitArrange dictionary where model is saved as a value |
---|
| 407 | :param pars: the list of parameters to fit |
---|
| 408 | :param constraints: list of |
---|
| 409 | tuple (name of parameter, value of parameters) |
---|
| 410 | the value of parameter must be a string to constraint 2 different |
---|
| 411 | parameters. |
---|
| 412 | Example: |
---|
| 413 | we want to fit 2 model M1 and M2 both have parameters A and B. |
---|
| 414 | constraints can be ``constraints = [(M1.A, M2.B+2), (M1.B= M2.A *5),...,]`` |
---|
| 415 | |
---|
| 416 | |
---|
| 417 | :note: pars must contains only name of existing model's parameters |
---|
| 418 | |
---|
| 419 | """ |
---|
[8d074d9] | 420 | if not pars: |
---|
| 421 | raise ValueError("no fitting parameters") |
---|
| 422 | |
---|
| 423 | if model is None: |
---|
| 424 | raise ValueError("no model to fit") |
---|
| 425 | |
---|
[51f14603] | 426 | if not issubclass(model.__class__, Model): |
---|
[95d58d3] | 427 | model = Model(model, data) |
---|
| 428 | |
---|
| 429 | sasmodel = model.model |
---|
[8d074d9] | 430 | available_parameters = sasmodel.getParamList() |
---|
| 431 | for p in pars: |
---|
| 432 | if p not in available_parameters: |
---|
| 433 | raise ValueError("parameter %s not available in model %s; use one of [%s] instead" |
---|
| 434 | %(p, sasmodel.name, ", ".join(available_parameters))) |
---|
| 435 | |
---|
| 436 | if id not in self.fit_arrange_dict: |
---|
| 437 | self.fit_arrange_dict[id] = FitArrange() |
---|
| 438 | |
---|
| 439 | self.fit_arrange_dict[id].set_model(model) |
---|
| 440 | self.fit_arrange_dict[id].pars = pars |
---|
| 441 | self.fit_arrange_dict[id].vals = [sasmodel.getParam(name) for name in pars] |
---|
| 442 | self.fit_arrange_dict[id].constraints = constraints |
---|
| 443 | |
---|
[51f14603] | 444 | def set_data(self, data, id, smearer=None, qmin=None, qmax=None): |
---|
| 445 | """ |
---|
| 446 | Receives plottable, creates a list of data to fit,set data |
---|
| 447 | in a FitArrange object and adds that object in a dictionary |
---|
| 448 | with key id. |
---|
| 449 | |
---|
| 450 | :param data: data added |
---|
| 451 | :param id: unique key corresponding to a fitArrange object with data |
---|
| 452 | """ |
---|
| 453 | if data.__class__.__name__ == 'Data2D': |
---|
[fd5ac0d] | 454 | fitdata = FitData2D(sas_data2d=data, data=data.data, |
---|
[51f14603] | 455 | err_data=data.err_data) |
---|
| 456 | else: |
---|
| 457 | fitdata = FitData1D(x=data.x, y=data.y, |
---|
| 458 | dx=data.dx, dy=data.dy, smearer=smearer) |
---|
[fd5ac0d] | 459 | fitdata.sas_data = data |
---|
[51f14603] | 460 | |
---|
| 461 | fitdata.set_fit_range(qmin=qmin, qmax=qmax) |
---|
| 462 | #A fitArrange is already created but contains model only at id |
---|
| 463 | if id in self.fit_arrange_dict: |
---|
| 464 | self.fit_arrange_dict[id].add_data(fitdata) |
---|
| 465 | else: |
---|
| 466 | #no fitArrange object has been create with this id |
---|
| 467 | fitproblem = FitArrange() |
---|
| 468 | fitproblem.add_data(fitdata) |
---|
| 469 | self.fit_arrange_dict[id] = fitproblem |
---|
| 470 | |
---|
| 471 | def get_model(self, id): |
---|
| 472 | """ |
---|
| 473 | :param id: id is key in the dictionary containing the model to return |
---|
| 474 | |
---|
| 475 | :return: a model at this id or None if no FitArrange element was |
---|
| 476 | created with this id |
---|
| 477 | """ |
---|
| 478 | if id in self.fit_arrange_dict: |
---|
| 479 | return self.fit_arrange_dict[id].get_model() |
---|
| 480 | else: |
---|
| 481 | return None |
---|
| 482 | |
---|
| 483 | def remove_fit_problem(self, id): |
---|
| 484 | """remove fitarrange in id""" |
---|
| 485 | if id in self.fit_arrange_dict: |
---|
| 486 | del self.fit_arrange_dict[id] |
---|
| 487 | |
---|
| 488 | def select_problem_for_fit(self, id, value): |
---|
| 489 | """ |
---|
| 490 | select a couple of model and data at the id position in dictionary |
---|
| 491 | and set in self.selected value to value |
---|
| 492 | |
---|
| 493 | :param value: the value to allow fitting. |
---|
| 494 | can only have the value one or zero |
---|
| 495 | """ |
---|
| 496 | if id in self.fit_arrange_dict: |
---|
| 497 | self.fit_arrange_dict[id].set_to_fit(value) |
---|
| 498 | |
---|
| 499 | def get_problem_to_fit(self, id): |
---|
| 500 | """ |
---|
| 501 | return the self.selected value of the fit problem of id |
---|
| 502 | |
---|
| 503 | :param id: the id of the problem |
---|
| 504 | """ |
---|
| 505 | if id in self.fit_arrange_dict: |
---|
| 506 | self.fit_arrange_dict[id].get_to_fit() |
---|
| 507 | |
---|
| 508 | |
---|
| 509 | class FitArrange: |
---|
| 510 | def __init__(self): |
---|
| 511 | """ |
---|
| 512 | Class FitArrange contains a set of data for a given model |
---|
| 513 | to perform the Fit.FitArrange must contain exactly one model |
---|
| 514 | and at least one data for the fit to be performed. |
---|
| 515 | |
---|
| 516 | model: the model selected by the user |
---|
| 517 | Ldata: a list of data what the user wants to fit |
---|
| 518 | |
---|
| 519 | """ |
---|
| 520 | self.model = None |
---|
| 521 | self.data_list = [] |
---|
| 522 | self.pars = [] |
---|
| 523 | self.vals = [] |
---|
| 524 | self.selected = 0 |
---|
[8d074d9] | 525 | |
---|
[51f14603] | 526 | def set_model(self, model): |
---|
| 527 | """ |
---|
| 528 | set_model save a copy of the model |
---|
| 529 | |
---|
| 530 | :param model: the model being set |
---|
| 531 | """ |
---|
| 532 | self.model = model |
---|
| 533 | |
---|
| 534 | def add_data(self, data): |
---|
| 535 | """ |
---|
| 536 | add_data fill a self.data_list with data to fit |
---|
| 537 | |
---|
| 538 | :param data: Data to add in the list |
---|
| 539 | """ |
---|
| 540 | if not data in self.data_list: |
---|
| 541 | self.data_list.append(data) |
---|
| 542 | |
---|
| 543 | def get_model(self): |
---|
| 544 | """ |
---|
| 545 | :return: saved model |
---|
| 546 | """ |
---|
| 547 | return self.model |
---|
| 548 | |
---|
| 549 | def get_data(self): |
---|
| 550 | """ |
---|
| 551 | :return: list of data data_list |
---|
| 552 | """ |
---|
| 553 | return self.data_list[0] |
---|
| 554 | |
---|
| 555 | def remove_data(self, data): |
---|
| 556 | """ |
---|
| 557 | Remove one element from the list |
---|
| 558 | |
---|
| 559 | :param data: Data to remove from data_list |
---|
| 560 | """ |
---|
| 561 | if data in self.data_list: |
---|
| 562 | self.data_list.remove(data) |
---|
| 563 | |
---|
| 564 | def set_to_fit(self, value=0): |
---|
| 565 | """ |
---|
| 566 | set self.selected to 0 or 1 for other values raise an exception |
---|
| 567 | |
---|
| 568 | :param value: integer between 0 or 1 |
---|
| 569 | """ |
---|
| 570 | self.selected = value |
---|
| 571 | |
---|
| 572 | def get_to_fit(self): |
---|
| 573 | """ |
---|
| 574 | return self.selected value |
---|
| 575 | """ |
---|
| 576 | return self.selected |
---|
[8d074d9] | 577 | |
---|
[51f14603] | 578 | class FResult(object): |
---|
| 579 | """ |
---|
| 580 | Storing fit result |
---|
| 581 | """ |
---|
| 582 | def __init__(self, model=None, param_list=None, data=None): |
---|
| 583 | self.calls = None |
---|
| 584 | self.fitness = None |
---|
| 585 | self.chisqr = None |
---|
| 586 | self.pvec = [] |
---|
| 587 | self.cov = [] |
---|
| 588 | self.info = None |
---|
| 589 | self.mesg = None |
---|
| 590 | self.success = None |
---|
| 591 | self.stderr = None |
---|
| 592 | self.residuals = [] |
---|
| 593 | self.index = [] |
---|
| 594 | self.model = model |
---|
| 595 | self.data = data |
---|
| 596 | self.theory = [] |
---|
| 597 | self.param_list = param_list |
---|
| 598 | self.iterations = 0 |
---|
| 599 | self.inputs = [] |
---|
| 600 | self.fitter_id = None |
---|
| 601 | if self.model is not None and self.data is not None: |
---|
| 602 | self.inputs = [(self.model, self.data)] |
---|
| 603 | |
---|
| 604 | def set_model(self, model): |
---|
| 605 | """ |
---|
| 606 | """ |
---|
| 607 | self.model = model |
---|
| 608 | |
---|
| 609 | def set_fitness(self, fitness): |
---|
| 610 | """ |
---|
| 611 | """ |
---|
| 612 | self.fitness = fitness |
---|
| 613 | |
---|
| 614 | def __str__(self): |
---|
| 615 | """ |
---|
| 616 | """ |
---|
| 617 | if self.pvec == None and self.model is None and self.param_list is None: |
---|
| 618 | return "No results" |
---|
[6fe5100] | 619 | |
---|
[95d58d3] | 620 | sasmodel = self.model.model |
---|
| 621 | pars = enumerate(sasmodel.getParamList()) |
---|
[6fe5100] | 622 | msg1 = "[Iteration #: %s ]" % self.iterations |
---|
| 623 | msg3 = "=== goodness of fit: %s ===" % (str(self.fitness)) |
---|
[95d58d3] | 624 | msg2 = ["P%-3d %s......|.....%s" % (i, v, sasmodel.getParam(v)) |
---|
[6fe5100] | 625 | for i,v in pars if v in self.param_list] |
---|
| 626 | msg = [msg1, msg3] + msg2 |
---|
| 627 | return "\n".join(msg) |
---|
[51f14603] | 628 | |
---|
| 629 | def print_summary(self): |
---|
| 630 | """ |
---|
| 631 | """ |
---|
[95d58d3] | 632 | print str(self) |
---|