[792db7d5] | 1 | """ |
---|
| 2 | @organization: ParkFitting module contains SansParameter,Model,Data |
---|
| 3 | FitArrange, ParkFit,Parameter classes.All listed classes work together to perform a |
---|
| 4 | simple fit with park optimizer. |
---|
| 5 | """ |
---|
[7705306] | 6 | import time |
---|
| 7 | import numpy |
---|
[792db7d5] | 8 | |
---|
[7705306] | 9 | import park |
---|
| 10 | from park import fit,fitresult |
---|
| 11 | from park import assembly |
---|
[cf3b781] | 12 | from park.fitmc import FitSimplex, FitMC |
---|
[7705306] | 13 | |
---|
| 14 | from sans.guitools.plottables import Data1D |
---|
| 15 | from Loader import Load |
---|
[d4b0687] | 16 | from AbstractFitEngine import FitEngine, Parameter, FitArrange |
---|
[7705306] | 17 | class SansParameter(park.Parameter): |
---|
| 18 | """ |
---|
[792db7d5] | 19 | SANS model parameters for use in the PARK fitting service. |
---|
| 20 | The parameter attribute value is redirected to the underlying |
---|
| 21 | parameter value in the SANS model. |
---|
[7705306] | 22 | """ |
---|
| 23 | def __init__(self, name, model): |
---|
| 24 | self._model, self._name = model,name |
---|
[9e85792] | 25 | self.set(model.getParam(name)) |
---|
[792db7d5] | 26 | |
---|
[7705306] | 27 | def _getvalue(self): return self._model.getParam(self.name) |
---|
[792db7d5] | 28 | |
---|
[9e85792] | 29 | def _setvalue(self,value): |
---|
| 30 | self._model.setParam(self.name, value) |
---|
[792db7d5] | 31 | |
---|
[7705306] | 32 | value = property(_getvalue,_setvalue) |
---|
[792db7d5] | 33 | |
---|
[7705306] | 34 | def _getrange(self): |
---|
| 35 | lo,hi = self._model.details[self.name][1:] |
---|
| 36 | if lo is None: lo = -numpy.inf |
---|
| 37 | if hi is None: hi = numpy.inf |
---|
| 38 | return lo,hi |
---|
[792db7d5] | 39 | |
---|
[7705306] | 40 | def _setrange(self,r): |
---|
| 41 | self._model.details[self.name][1:] = r |
---|
| 42 | range = property(_getrange,_setrange) |
---|
| 43 | |
---|
[792db7d5] | 44 | |
---|
[7705306] | 45 | class Model(object): |
---|
| 46 | """ |
---|
| 47 | PARK wrapper for SANS models. |
---|
| 48 | """ |
---|
| 49 | def __init__(self, sans_model): |
---|
| 50 | self.model = sans_model |
---|
| 51 | sansp = sans_model.getParamList() |
---|
| 52 | parkp = [SansParameter(p,sans_model) for p in sansp] |
---|
| 53 | self.parameterset = park.ParameterSet(sans_model.name,pars=parkp) |
---|
[792db7d5] | 54 | |
---|
[7705306] | 55 | def eval(self,x): |
---|
| 56 | return self.model.run(x) |
---|
| 57 | |
---|
| 58 | class Data(object): |
---|
| 59 | """ Wrapper class for SANS data """ |
---|
[792db7d5] | 60 | def __init__(self,x=None,y=None,dy=None,dx=None,sans_data=None): |
---|
| 61 | if not sans_data==None: |
---|
| 62 | self.x= sans_data.x |
---|
| 63 | self.y= sans_data.y |
---|
| 64 | self.dx= sans_data.dx |
---|
| 65 | self.dy= sans_data.dy |
---|
| 66 | else: |
---|
| 67 | if x!=None and y!=None and dy!=None: |
---|
| 68 | self.x=x |
---|
| 69 | self.y=y |
---|
| 70 | self.dx=dx |
---|
| 71 | self.dy=dy |
---|
| 72 | else: |
---|
| 73 | raise ValueError,\ |
---|
| 74 | "Data is missing x, y or dy, impossible to compute residuals later on" |
---|
[7705306] | 75 | self.qmin=None |
---|
| 76 | self.qmax=None |
---|
| 77 | |
---|
| 78 | def setFitRange(self,mini=None,maxi=None): |
---|
| 79 | """ to set the fit range""" |
---|
| 80 | self.qmin=mini |
---|
| 81 | self.qmax=maxi |
---|
| 82 | |
---|
| 83 | def residuals(self, fn): |
---|
[792db7d5] | 84 | """ @param fn: function that return model value |
---|
| 85 | @return residuals |
---|
| 86 | """ |
---|
[7705306] | 87 | x,y,dy = [numpy.asarray(v) for v in (self.x,self.y,self.dy)] |
---|
| 88 | if self.qmin==None and self.qmax==None: |
---|
[cf3b781] | 89 | self.fx = fn(x) |
---|
[7705306] | 90 | return (y - fn(x))/dy |
---|
| 91 | |
---|
| 92 | else: |
---|
[cf3b781] | 93 | self.fx = fn(x[idx]) |
---|
[7705306] | 94 | idx = x>=self.qmin & x <= self.qmax |
---|
| 95 | return (y[idx] - fn(x[idx]))/dy[idx] |
---|
| 96 | |
---|
| 97 | |
---|
| 98 | def residuals_deriv(self, model, pars=[]): |
---|
[792db7d5] | 99 | """ |
---|
| 100 | @return residuals derivatives . |
---|
| 101 | @note: in this case just return empty array |
---|
| 102 | """ |
---|
[7705306] | 103 | return [] |
---|
[d4b0687] | 104 | |
---|
[792db7d5] | 105 | |
---|
[4c718654] | 106 | class ParkFit(FitEngine): |
---|
[7705306] | 107 | """ |
---|
[792db7d5] | 108 | ParkFit performs the Fit.This class can be used as follow: |
---|
| 109 | #Do the fit Park |
---|
| 110 | create an engine: engine = ParkFit() |
---|
| 111 | Use data must be of type plottable |
---|
| 112 | Use a sans model |
---|
| 113 | |
---|
| 114 | Add data with a dictionnary of FitArrangeList where Uid is a key and data |
---|
| 115 | is saved in FitArrange object. |
---|
| 116 | engine.set_data(data,Uid) |
---|
| 117 | |
---|
| 118 | Set model parameter "M1"= model.name add {model.parameter.name:value}. |
---|
| 119 | @note: Set_param() if used must always preceded set_model() |
---|
| 120 | for the fit to be performed. |
---|
| 121 | engine.set_param( model,"M1", {'A':2,'B':4}) |
---|
| 122 | |
---|
| 123 | Add model with a dictionnary of FitArrangeList{} where Uid is a key and model |
---|
| 124 | is save in FitArrange object. |
---|
| 125 | engine.set_model(model,Uid) |
---|
| 126 | |
---|
| 127 | engine.fit return chisqr,[model.parameter 1,2,..],[[err1....][..err2...]] |
---|
| 128 | chisqr1, out1, cov1=engine.fit({model.parameter.name:value},qmin,qmax) |
---|
| 129 | @note: {model.parameter.name:value} is ignored in fit function since |
---|
| 130 | the user should make sure to call set_param himself. |
---|
[7705306] | 131 | """ |
---|
| 132 | def __init__(self,data=[]): |
---|
[792db7d5] | 133 | """ |
---|
| 134 | Creates a dictionary (self.fitArrangeList={})of FitArrange elements |
---|
| 135 | with Uid as keys |
---|
| 136 | """ |
---|
[7705306] | 137 | self.fitArrangeList={} |
---|
[792db7d5] | 138 | |
---|
[4dd63eb] | 139 | def createProblem(self): |
---|
[7705306] | 140 | """ |
---|
[792db7d5] | 141 | Extract sansmodel and sansdata from self.FitArrangelist ={Uid:FitArrange} |
---|
| 142 | Create parkmodel and park data ,form a list couple of parkmodel and parkdata |
---|
| 143 | create an assembly self.problem= park.Assembly([(parkmodel,parkdata)]) |
---|
[7705306] | 144 | """ |
---|
| 145 | mylist=[] |
---|
[9e85792] | 146 | listmodel=[] |
---|
[7705306] | 147 | for k,value in self.fitArrangeList.iteritems(): |
---|
[9e85792] | 148 | sansmodel=value.get_model() |
---|
[792db7d5] | 149 | #wrap sans model |
---|
[9e85792] | 150 | parkmodel = Model(sansmodel) |
---|
| 151 | for p in parkmodel.parameterset: |
---|
| 152 | if p.isfixed(): |
---|
| 153 | p.set([-numpy.inf,numpy.inf]) |
---|
| 154 | |
---|
[7705306] | 155 | Ldata=value.get_data() |
---|
[d4b0687] | 156 | x,y,dy=self._concatenateData(Ldata) |
---|
[792db7d5] | 157 | #wrap sansdata |
---|
[d4b0687] | 158 | parkdata=Data(x,y,dy,None) |
---|
[792db7d5] | 159 | couple=(parkmodel,parkdata) |
---|
[7705306] | 160 | mylist.append(couple) |
---|
[792db7d5] | 161 | |
---|
[cf3b781] | 162 | self.problem = park.Assembly(mylist) |
---|
[792db7d5] | 163 | |
---|
[7705306] | 164 | |
---|
[4dd63eb] | 165 | def fit(self, qmin=None, qmax=None): |
---|
[7705306] | 166 | """ |
---|
[792db7d5] | 167 | Performs fit with park.fit module.It can perform fit with one model |
---|
| 168 | and a set of data, more than two fit of one model and sets of data or |
---|
| 169 | fit with more than two model associated with their set of data and constraints |
---|
| 170 | |
---|
| 171 | |
---|
| 172 | @param pars: Dictionary of parameter names for the model and their values. |
---|
| 173 | @param qmin: The minimum value of data's range to be fit |
---|
| 174 | @param qmax: The maximum value of data's range to be fit |
---|
| 175 | @note:all parameter are ignored most of the time.Are just there to keep ScipyFit |
---|
| 176 | and ParkFit interface the same. |
---|
| 177 | @return result.fitness: Value of the goodness of fit metric |
---|
| 178 | @return result.pvec: list of parameter with the best value found during fitting |
---|
| 179 | @return result.cov: Covariance matrix |
---|
[7705306] | 180 | """ |
---|
[cf3b781] | 181 | |
---|
[792db7d5] | 182 | |
---|
[4dd63eb] | 183 | self.createProblem() |
---|
[cf3b781] | 184 | pars=self.problem.fit_parameters() |
---|
| 185 | self.problem.eval() |
---|
[792db7d5] | 186 | |
---|
[cf3b781] | 187 | localfit = FitSimplex() |
---|
| 188 | localfit.ftol = 1e-8 |
---|
| 189 | fitter = FitMC(localfit=localfit) |
---|
| 190 | |
---|
| 191 | result = fit.fit(self.problem, |
---|
| 192 | fitter=fitter, |
---|
| 193 | handler= fitresult.ConsoleUpdate(improvement_delta=0.1)) |
---|
[df58d26f] | 194 | |
---|
[792db7d5] | 195 | return result.fitness,result.pvec,result.cov |
---|
[7705306] | 196 | |
---|
[d4b0687] | 197 | |
---|