Added new boost headers for gamma distribution as well.

bob/core/include/bob.core/boost/exponential_distribution.hpp

@@ -24,7 +24,6 @@
 #include <boost/random/detail/config.hpp>
 #include <boost/random/uniform_01.hpp>
 
-
 #include <bob.core/boost/operators.hpp>
 
 namespace bob {

bob/core/include/bob.core/boost/gamma_distribution.hpp

+/* boost random/gamma_distribution.hpp header file
+ *
+ * Copyright Jens Maurer 2002
+ * Copyright Steven Watanabe 2010
+ * Distributed under the Boost Software License, Version 1.0. (See
+ * accompanying file LICENSE_1_0.txt or copy at
+ * http://www.boost.org/LICENSE_1_0.txt)
+ *
+ * See http://www.boost.org for most recent version including documentation.
+ *
+ * $Id$
+ *
+ */
+
+#ifndef BOB_BOOST_RANDOM_GAMMA_DISTRIBUTION_HPP
+#define BOB_BOOST_RANDOM_GAMMA_DISTRIBUTION_HPP
+
+#include <boost/config/no_tr1/cmath.hpp>
+#include <istream>
+#include <iosfwd>
+#include <boost/assert.hpp>
+#include <boost/limits.hpp>
+#include <boost/static_assert.hpp>
+#include <boost/random/detail/config.hpp>
+#include <boost/random/uniform_01.hpp>
+
+#include <bob.core/boost/exponential_distribution.hpp>
+
+namespace bob {
+namespace core {
+namespace random {
+
+// The algorithm is taken from Knuth
+
+/**
+ * The gamma distribution is a continuous distribution with two
+ * parameters alpha and beta.  It produces values > 0.
+ *
+ * It has
+ * \f$\displaystyle p(x) = x^{\alpha-1}\frac{e^{-x/\beta}}{\beta^\alpha\Gamma(\alpha)}\f$.
+ */
+template<class RealType = double>
+class gamma_distribution
+{
+public:
+    typedef RealType input_type;
+    typedef RealType result_type;
+
+    class param_type
+    {
+    public:
+        typedef gamma_distribution distribution_type;
+
+        /**
+         * Constructs a @c param_type object from the "alpha" and "beta"
+         * parameters.
+         *
+         * Requires: alpha > 0 && beta > 0
+         */
+        param_type(const RealType& alpha_arg = RealType(1.0),
+                   const RealType& beta_arg = RealType(1.0))
+          : _alpha(alpha_arg), _beta(beta_arg)
+        {
+        }
+
+        /** Returns the "alpha" parameter of the distribution. */
+        RealType alpha() const { return _alpha; }
+        /** Returns the "beta" parameter of the distribution. */
+        RealType beta() const { return _beta; }
+
+#ifndef BOOST_RANDOM_NO_STREAM_OPERATORS
+        /** Writes the parameters to a @c std::ostream. */
+        template<class CharT, class Traits>
+        friend std::basic_ostream<CharT, Traits>&
+        operator<<(std::basic_ostream<CharT, Traits>& os,
+                   const param_type& parm)
+        {
+            os << parm._alpha << ' ' << parm._beta;
+            return os;
+        }
+
+        /** Reads the parameters from a @c std::istream. */
+        template<class CharT, class Traits>
+        friend std::basic_istream<CharT, Traits>&
+        operator>>(std::basic_istream<CharT, Traits>& is, param_type& parm)
+        {
+            is >> parm._alpha >> std::ws >> parm._beta;
+            return is;
+        }
+#endif
+
+        /** Returns true if the two sets of parameters are the same. */
+        friend bool operator==(const param_type& lhs, const param_type& rhs)
+        {
+            return lhs._alpha == rhs._alpha && lhs._beta == rhs._beta;
+        }
+        /** Returns true if the two sets fo parameters are different. */
+        friend bool operator!=(const param_type& lhs, const param_type& rhs)
+        {
+            return !(lhs == rhs);
+        }
+    private:
+        RealType _alpha;
+        RealType _beta;
+    };
+
+#ifndef BOOST_NO_LIMITS_COMPILE_TIME_CONSTANTS
+    BOOST_STATIC_ASSERT(!std::numeric_limits<RealType>::is_integer);
+#endif
+
+    /**
+     * Creates a new gamma_distribution with parameters "alpha" and "beta".
+     *
+     * Requires: alpha > 0 && beta > 0
+     */
+    explicit gamma_distribution(const result_type& alpha_arg = result_type(1.0),
+                                const result_type& beta_arg = result_type(1.0))
+      : _exp(result_type(1)), _alpha(alpha_arg), _beta(beta_arg)
+    {
+        BOOST_ASSERT(_alpha > result_type(0));
+        BOOST_ASSERT(_beta > result_type(0));
+        init();
+    }
+
+    /** Constructs a @c gamma_distribution from its parameters. */
+    explicit gamma_distribution(const param_type& parm)
+      : _exp(result_type(1)), _alpha(parm.alpha()), _beta(parm.beta())
+    {
+        init();
+    }
+
+    // compiler-generated copy ctor and assignment operator are fine
+
+    /** Returns the "alpha" paramter of the distribution. */
+    RealType alpha() const { return _alpha; }
+    /** Returns the "beta" parameter of the distribution. */
+    RealType beta() const { return _beta; }
+    /** Returns the smallest value that the distribution can produce. */
+    RealType min BOOST_PREVENT_MACRO_SUBSTITUTION () const { return 0; }
+    /* Returns the largest value that the distribution can produce. */
+    RealType max BOOST_PREVENT_MACRO_SUBSTITUTION () const
+    { return (std::numeric_limits<RealType>::infinity)(); }
+
+    /** Returns the parameters of the distribution. */
+    param_type param() const { return param_type(_alpha, _beta); }
+    /** Sets the parameters of the distribution. */
+    void param(const param_type& parm)
+    {
+        _alpha = parm.alpha();
+        _beta = parm.beta();
+        init();
+    }
+
+    /**
+     * Effects: Subsequent uses of the distribution do not depend
+     * on values produced by any engine prior to invoking reset.
+     */
+    void reset() { _exp.reset(); }
+
+    /**
+     * Returns a random variate distributed according to
+     * the gamma distribution.
+     */
+    template<class Engine>
+    result_type operator()(Engine& eng)
+    {
+#ifndef BOOST_NO_STDC_NAMESPACE
+        // allow for Koenig lookup
+        using std::tan; using std::sqrt; using std::exp; using std::log;
+        using std::pow;
+#endif
+        if(_alpha == result_type(1)) {
+            return _exp(eng) * _beta;
+        } else if(_alpha > result_type(1)) {
+            // Can we have a boost::mathconst please?
+            const result_type pi = result_type(3.14159265358979323846);
+            for(;;) {
+                result_type y = tan(pi * boost::uniform_01<RealType>()(eng));
+                result_type x = sqrt(result_type(2)*_alpha-result_type(1))*y
+                    + _alpha-result_type(1);
+                if(x <= result_type(0))
+                    continue;
+                if(boost::uniform_01<RealType>()(eng) >
+                    (result_type(1)+y*y) * exp((_alpha-result_type(1))
+                                               *log(x/(_alpha-result_type(1)))
+                                               - sqrt(result_type(2)*_alpha
+                                                      -result_type(1))*y))
+                    continue;
+                return x * _beta;
+            }
+        } else /* alpha < 1.0 */ {
+            for(;;) {
+                result_type u = boost::uniform_01<RealType>()(eng);
+                result_type y = _exp(eng);
+                result_type x, q;
+                if(u < _p) {
+                    x = exp(-y/_alpha);
+                    q = _p*exp(-x);
+                } else {
+                    x = result_type(1)+y;
+                    q = _p + (result_type(1)-_p) * pow(x,_alpha-result_type(1));
+                }
+                if(u >= q)
+                    continue;
+                return x * _beta;
+            }
+        }
+    }
+
+    template<class URNG>
+    RealType operator()(URNG& urng, const param_type& parm) const
+    {
+        return gamma_distribution(parm)(urng);
+    }
+
+#ifndef BOOST_RANDOM_NO_STREAM_OPERATORS
+    /** Writes a @c gamma_distribution to a @c std::ostream. */
+    template<class CharT, class Traits>
+    friend std::basic_ostream<CharT,Traits>&
+    operator<<(std::basic_ostream<CharT,Traits>& os,
+               const gamma_distribution& gd)
+    {
+        os << gd.param();
+        return os;
+    }
+
+    /** Reads a @c gamma_distribution from a @c std::istream. */
+    template<class CharT, class Traits>
+    friend std::basic_istream<CharT,Traits>&
+    operator>>(std::basic_istream<CharT,Traits>& is, gamma_distribution& gd)
+    {
+        gd.read(is);
+        return is;
+    }
+#endif
+
+    /**
+     * Returns true if the two distributions will produce identical
+     * sequences of random variates given equal generators.
+     */
+    friend bool operator==(const gamma_distribution& lhs,
+                           const gamma_distribution& rhs)
+    {
+        return lhs._alpha == rhs._alpha
+            && lhs._beta == rhs._beta
+            && lhs._exp == rhs._exp;
+    }
+
+    /**
+     * Returns true if the two distributions can produce different
+     * sequences of random variates, given equal generators.
+     */
+    friend bool operator!=(const gamma_distribution& lhs,
+                           const gamma_distribution& rhs)
+    {
+        return !(lhs == rhs);
+    }
+
+private:
+    /// \cond hide_private_members
+
+    template<class CharT, class Traits>
+    void read(std::basic_istream<CharT, Traits>& is)
+    {
+        param_type parm;
+        if(is >> parm) {
+            param(parm);
+        }
+    }
+
+    void init()
+    {
+#ifndef BOOST_NO_STDC_NAMESPACE
+        // allow for Koenig lookup
+        using std::exp;
+#endif
+        _p = exp(result_type(1)) / (_alpha + exp(result_type(1)));
+    }
+    /// \endcond
+
+    bob::core::random::exponential_distribution<RealType> _exp;
+    result_type _alpha;
+    result_type _beta;
+    // some data precomputed from the parameters
+    result_type _p;
+};
+
+} } } // namespaces
+
+#endif // BOB_BOOST_RANDOM_GAMMA_DISTRIBUTION_HPP

bob/core/include/bob.core/random.h

@@ -21,12 +21,16 @@ namespace bob { namespace core { namespace random {
   template <class RealType=double>
     using lognormal_distribution = boost::random::lognormal_distribution<RealType>;
 
+  template<class RealType=double>
+    using gamma_distribution = boost::random::gamma_distribution<RealType>;
+
   template<class IntType=int, class RealType=double>
     using binomial_distribution = boost::random::binomial_distribution<IntType, RealType>;
 
-  template<class IntType, class WeightType>
+  template<class IntType=int, class WeightType=double>
     using discrete_distribution = boost::random::discrete_distribution<IntType, WeightType>;
 
+
 } } } // namespaces
 
 #else
@@ -37,6 +41,7 @@ namespace bob { namespace core { namespace random {
 #include <bob.core/boost/lognormal_distribution.hpp>
 #include <bob.core/boost/binomial_distribution.hpp>
 #include <bob.core/boost/discrete_distribution.hpp>
+#include <bob.core/boost/gamma_distribution.hpp>
 
 #endif // BOOST VERSION
 

bob/core/random/gamma.cpp

@@ -53,7 +53,7 @@ template <typename T>
 boost::shared_ptr<void> make_gamma(PyObject* alpha) {
   T calpha = 1.;
   if (alpha) calpha = PyBlitzArrayCxx_AsCScalar<T>(alpha);
-  return boost::make_shared<boost::gamma_distribution<T>>(calpha);
+  return boost::make_shared<bob::core::random::gamma_distribution<T>>(calpha);
 }
 
 PyObject* PyBoostGamma_SimpleNew (int type_num, PyObject* alpha) {
@@ -130,7 +130,7 @@ int PyBoostGamma_Converter(PyObject* o, PyBoostGammaObject** a) {
 }
 
 template <typename T> PyObject* get_alpha(PyBoostGammaObject* self) {
-  return PyBlitzArrayCxx_FromCScalar(boost::static_pointer_cast<boost::gamma_distribution<T>>(self->distro)->alpha());
+  return PyBlitzArrayCxx_FromCScalar(boost::static_pointer_cast<bob::core::random::gamma_distribution<T>>(self->distro)->alpha());
 }
 
 /**
@@ -156,7 +156,7 @@ static PyObject* PyBoostGamma_GetDtype(PyBoostGammaObject* self) {
 }
 
 template <typename T> PyObject* reset(PyBoostGammaObject* self) {
-  boost::static_pointer_cast<boost::gamma_distribution<T>>(self->distro)->reset();
+  boost::static_pointer_cast<bob::core::random::gamma_distribution<T>>(self->distro)->reset();
   Py_RETURN_NONE;
 }
 
@@ -177,7 +177,7 @@ static PyObject* PyBoostGamma_Reset(PyBoostGammaObject* self) {
 }
 
 template <typename T> PyObject* call(PyBoostGammaObject* self, PyBoostMt19937Object* rng) {
-  typedef boost::gamma_distribution<T> distro_t;
+  typedef bob::core::random::gamma_distribution<T> distro_t;
   return PyBlitzArrayCxx_FromCScalar((*boost::static_pointer_cast<distro_t>(self->distro))(*rng->rng));
 }