LSSTApplications  18.1.0
LSSTDataManagementBasePackage
TruncatedGaussian.h
Go to the documentation of this file.
1 // -*- lsst-c++ -*-
2 /*
3  * LSST Data Management System
4  * Copyright 2008-2013 LSST Corporation.
5  *
6  * This product includes software developed by the
7  * LSST Project (http://www.lsst.org/).
8  *
9  * This program is free software: you can redistribute it and/or modify
10  * it under the terms of the GNU General Public License as published by
11  * the Free Software Foundation, either version 3 of the License, or
12  * (at your option) any later version.
13  *
14  * This program is distributed in the hope that it will be useful,
15  * but WITHOUT ANY WARRANTY; without even the implied warranty of
16  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
17  * GNU General Public License for more details.
18  *
19  * You should have received a copy of the LSST License Statement and
20  * the GNU General Public License along with this program. If not,
21  * see <http://www.lsstcorp.org/LegalNotices/>.
22  */
23 
24 #ifndef LSST_MEAS_MODELFIT_TruncatedGaussian_h_INCLUDED
25 #define LSST_MEAS_MODELFIT_TruncatedGaussian_h_INCLUDED
26 
27 #include "Eigen/Core"
28 #include "ndarray.h"
29 
30 #include "lsst/base.h"
31 #include "lsst/afw/math/Random.h"
33 
34 // TODO: we should really integrate this with Mixture somehow
35 
36 namespace lsst { namespace meas { namespace modelfit {
37 
38 class TruncatedGaussianSampler;
39 class TruncatedGaussianEvaluator;
40 class TruncatedGaussianLogEvaluator;
41 
60 public:
61 
64 
85  };
86 
90 
125  static TruncatedGaussian fromSeriesParameters(Scalar q0, Vector const & gradient, Matrix const & hessian);
126 
154  static TruncatedGaussian fromStandardParameters(Vector const & mean, Matrix const & covariance);
155 
163  Sampler sample(SampleStrategy strategy) const;
164 
173  Sampler sample(Scalar minRejectionEfficiency=0.1) const;
174 
176  LogEvaluator evaluateLog() const;
177 
179  Evaluator evaluate() const;
180 
182  int getDim() const;
183 
191  Vector maximize() const;
192 
204 
211  Scalar getLogPeakAmplitude() const;
212 
226  Scalar getLogIntegral() const;
227 
229 
230 private:
231 
234 
235  class Impl;
236 
237  explicit TruncatedGaussian(PTR(Impl) impl) : _impl(impl) {}
238 
239  PTR(Impl) _impl;
240 };
241 
246 public:
247 
248  explicit TruncatedGaussianLogEvaluator(TruncatedGaussian const & parent);
249 
250  template <typename Derived>
251  Scalar operator()(Eigen::MatrixBase<Derived> const & alpha) const {
252  if ((alpha.array() < 0.0).any()) return std::numeric_limits<Scalar>::infinity();
253  _workspace = alpha - _mu;
254  return 0.5*(_rootH*_workspace).squaredNorm() + _norm;
255  }
256 
257  Scalar operator()(ndarray::Array<Scalar const,1,1> const & alpha) const;
258 
259  void operator()(
260  ndarray::Array<Scalar const,2,1> const & alpha,
261  ndarray::Array<Scalar,1,1> const & output
262  ) const;
263 
264 protected:
269 };
270 
275 public:
276 
278  _internal(parent)
279  {}
280 
281  template <typename Derived>
282  Scalar operator()(Eigen::MatrixBase<Derived> const & alpha) const {
283  return std::exp(-_internal(alpha));
284  }
285 
286  Scalar operator()(ndarray::Array<Scalar const,1,1> const & alpha) const;
287 
288  void operator()(
289  ndarray::Array<Scalar const,2,1> const & alpha,
290  ndarray::Array<Scalar,1,1> const & output
291  ) const;
292 
293 private:
295 };
296 
301 public:
302 
303  explicit TruncatedGaussianSampler(
304  TruncatedGaussian const & parent,
306  );
307 
316  Scalar operator()(afw::math::Random & rng, ndarray::Array<Scalar,1,1> const & alpha) const;
317 
327  void operator()(
328  afw::math::Random & rng,
329  ndarray::Array<Scalar,2,1> const & alpha,
330  ndarray::Array<Scalar,1,1> const & weights,
331  bool multiplyWeights=false
332  ) const;
333 
334  ~TruncatedGaussianSampler(); // defined in .cc so it can see Impl's dtor
335 
336  class Impl; // public so we can inherit from it in the .cc file
337 
338 private:
339  PTR(Impl) _impl;
340 };
341 
343  return Sampler(*this, strategy);
344 }
345 
346 inline TruncatedGaussian::Sampler TruncatedGaussian::sample(Scalar minRejectionEfficiency) const {
347  return Sampler(
348  *this,
349  (getUntruncatedFraction() < minRejectionEfficiency) ? ALIGN_AND_WEIGHT : DIRECT_WITH_REJECTION
350  );
351 }
352 
354 
356 
357 }}} // namespace lsst::meas::modelfit
358 
359 #endif // !LSST_MEAS_MODELFIT_TruncatedGaussian_h_INCLUDED
static TruncatedGaussian fromStandardParameters(Vector const &mean, Matrix const &covariance)
Create from the "standard" mean and covariance parameters of the normal distribution.
TruncatedGaussianLogEvaluator LogEvaluator
Scalar getLogPeakAmplitude() const
Return the -log of the peak amplitude of the untruncated function.
Basic LSST definitions.
#define PTR(...)
Definition: base.h:41
T exp(T... args)
TruncatedGaussianEvaluator(TruncatedGaussian const &parent)
Helper class for drawing samples from a TruncatedGaussian.
Sampler sample(SampleStrategy strategy) const
Create a Sampler object that uses the given strategy.
double Scalar
Typedefs to be used for probability and parameter values.
Definition: common.h:44
Scalar getUntruncatedFraction() const
Return the fraction of the Gaussian integral that was truncated by the bounds.
MatrixQ covariance
Definition: simpleShape.cc:152
Scalar operator()(Eigen::MatrixBase< Derived > const &alpha) const
A base class for image defects.
Eigen::Matrix< Scalar, Eigen::Dynamic, Eigen::Dynamic > Matrix
Typedefs to be used for probability and parameter values.
Definition: common.h:45
Represents a multidimensional Gaussian function truncated at zero.
Eigen::Matrix< Scalar, Eigen::Dynamic, 1 > Vector
Typedefs to be used for probability and parameter values.
Definition: common.h:46
Vector maximize() const
Return the location of the maximum of the truncated Gaussian.
SampleStrategy
Enum that describes different ways of sampling from a multidimensional TruncatedGaussian.
Helper class for evaluating the -log of a TruncatedGaussian.
Create a similar Gaussian with no x-y covariance, and importance sample by drawing from the independe...
LogEvaluator evaluateLog() const
Create a LogEvaluator object that can be used to efficiently evaluate the -log of the function...
Helper class for evaluating the -log of a TruncatedGaussian.
Evaluator evaluate() const
Create an Evaluator object that can be used to efficiently evaluate the function. ...
Draw from the untruncated Gaussian, and discard negative draws.
Scalar operator()(Eigen::MatrixBase< Derived > const &alpha) const
TruncatedGaussianEvaluator Evaluator
int getDim() const
Return the dimensionality of the function.
Scalar getLogIntegral() const
Return the -log of the integral of the truncated function.
static TruncatedGaussian fromSeriesParameters(Scalar q0, Vector const &gradient, Matrix const &hessian)
Create from the first and second logarithmic derivatives of the Gaussian.
A class that can be used to generate sequences of random numbers according to a number of different a...
Definition: Random.h:57