manolas
/
vcglib
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity
vcglib/vcg/space/deprecated_point4.h

396 lines
11 KiB
C
Raw Normal View History

Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
/****************************************************************************
* VCGLib o o *
* Visual and Computer Graphics Library o o *
* _ O _ *
* Copyright(C) 2004 \/)\/ *
* Visual Computing Lab /\/| *
* ISTI - Italian National Research Council | *
* \ *
* All rights reserved. *
* *
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
* This program is free software; you can redistribute it and/or modify *
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
* it under the terms of the GNU General Public License as published by *
* the Free Software Foundation; either version 2 of the License, or *
* (at your option) any later version. *
* *
* This program is distributed in the hope that it will be useful, *
* but WITHOUT ANY WARRANTY; without even the implied warranty of *
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the *
* GNU General Public License (http://www.gnu.org/licenses/gpl.txt) *
* for more details. *
* *
****************************************************************************/
/****************************************************************************
History
$Log: not supported by cvs2svn $
Revision 1.12 2006/06/21 11:06:16 ganovelli
changed return type of Zero() (to void)
Revision 1.11 2005/04/13 09:40:30 ponchio
Including math/bash.h
Revision 1.10 2005/03/18 16:34:42 fiorin
minor changes to comply gcc compiler
Revision 1.9 2005/01/21 18:02:11 ponchio
Removed dependence from matrix44 and changed VectProd
Revision 1.8 2005/01/12 11:25:02 ganovelli
added Dimension
Revision 1.7 2004/10/11 17:46:11 ganovelli
added definition of vector product (not implemented)
Revision 1.6 2004/05/10 11:16:19 ganovelli
include assert.h added
Revision 1.5 2004/03/31 10:09:58 cignoni
missing return value in zero()
Revision 1.4 2004/03/11 17:17:49 tarini
added commets (doxy), uniformed with new style, now using math::, ...
added HomoNormalize(), Zero()... remade StableDot() (hand made sort).
Revision 1.1 2004/02/10 01:11:28 cignoni
Edited Comments and GPL license
****************************************************************************/
#ifndef __VCGLIB_POINT4
#define __VCGLIB_POINT4
#include <assert.h>
#include <vcg/math/base.h>
namespace vcg {
/** \addtogroup space */
/*@{*/
/**
The templated class for representing a point in 4D space.
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
The class is templated over the ScalarType class that is used to represent coordinates.
All the usual operator (* + - ...) are defined.
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
*/
template <class T> class Point4
{
public:
/// The only data member. Hidden to user.
T _v[4];
public:
typedef T ScalarType;
enum {Dimension = 4};
//@{
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
/** @name Standard Constructors and Initializers
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
No casting operators have been introduced to avoid automatic unattended (and costly) conversion between different point types
**/
inline Point4 () { }
inline Point4 ( const T nx, const T ny, const T nz , const T nw )
{
_v[0] = nx; _v[1] = ny; _v[2] = nz; _v[3] = nw;
}
inline Point4 ( const T p[4] )
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
{
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
_v[0] = p[0]; _v[1]= p[1]; _v[2] = p[2]; _v[3]= p[3];
}
inline Point4 ( const Point4 & p )
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
{
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
_v[0]= p._v[0]; _v[1]= p._v[1]; _v[2]= p._v[2]; _v[3]= p._v[3];
}
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
inline void SetZero()
{
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
_v[0] = _v[1] = _v[2] = _v[3]= 0;
}
template <class Q>
/// importer from different Point4 types
inline void Import( const Point4<Q> & b )
{
_v[0] = T(b[0]); _v[1] = T(b[1]);
_v[2] = T(b[2]);
_v[3] = T(b[3]);
}
Deprecating the use of old dangerous linear algebra code. Please use Eigen! Now attempting to include lin_algebra will block your compilation...
2012-10-11 12:48:55 +02:00
template <class EigenVector>
inline void FromEigenVector( const EigenVector & b )
{
_v[0] = T(b[0]);
_v[1] = T(b[1]);
_v[2] = T(b[2]);
_v[3] = T(b[3]);
}
/// constructor that imports from different Point4 types
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
template <class Q>
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
static inline Point4 Construct( const Point4<Q> & b )
{
return Point4(T(b[0]),T(b[1]),T(b[2]),T(b[3]));
}
//@}
//@{
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
/** @name Data Access.
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
access to data is done by overloading of [] or explicit naming of coords (x,y,z,w)
**/
inline const T & operator [] ( const int i ) const
{
assert(i>=0 && i<4);
return _v[i];
}
inline T & operator [] ( const int i )
{
assert(i>=0 && i<4);
return _v[i];
}
inline T &X() {return _v[0];}
inline T &Y() {return _v[1];}
inline T &Z() {return _v[2];}
inline T &W() {return _v[3];}
inline T const * V() const
{
Added non-const pointer accessor.
2009-03-19 12:13:17 +01:00
return _v;
}
inline T * V()
{
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
return _v;
}
inline const T & V ( const int i ) const
{
assert(i>=0 && i<4);
return _v[i];
}
inline T & V ( const int i )
{
assert(i>=0 && i<4);
return _v[i];
}
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
/// Padding function: give a default 0 value to all the elements that are not in the [0..2] range.
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
/// Useful for managing in a consistent way object that could have point2 / point3 / point4
inline T Ext( const int i ) const
{
if(i>=0 && i<=3) return _v[i];
else return 0;
}
//@}
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
//@{
/** @name Linear operators and the likes
**/
inline Point4 operator + ( const Point4 & p) const
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
{
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
return Point4( _v[0]+p._v[0], _v[1]+p._v[1], _v[2]+p._v[2], _v[3]+p._v[3] );
}
inline Point4 operator - ( const Point4 & p) const
{
return Point4( _v[0]-p._v[0], _v[1]-p._v[1], _v[2]-p._v[2], _v[3]-p._v[3] );
}
inline Point4 operator * ( const T s ) const
{
return Point4( _v[0]*s, _v[1]*s, _v[2]*s, _v[3]*s );
}
inline Point4 operator / ( const T s ) const
{
return Point4( _v[0]/s, _v[1]/s, _v[2]/s, _v[3]/s );
}
inline Point4 & operator += ( const Point4 & p)
{
_v[0] += p._v[0]; _v[1] += p._v[1]; _v[2] += p._v[2]; _v[3] += p._v[3];
return *this;
}
inline Point4 & operator -= ( const Point4 & p )
{
_v[0] -= p._v[0]; _v[1] -= p._v[1]; _v[2] -= p._v[2]; _v[3] -= p._v[3];
return *this;
}
inline Point4 & operator *= ( const T s )
{
_v[0] *= s; _v[1] *= s; _v[2] *= s; _v[3] *= s;
return *this;
}
inline Point4 & operator /= ( const T s )
{
_v[0] /= s; _v[1] /= s; _v[2] /= s; _v[3] /= s;
return *this;
}
inline Point4 operator - () const
{
return Point4( -_v[0], -_v[1], -_v[2], -_v[3] );
}
inline Point4 VectProd ( const Point4 &x, const Point4 &z ) const
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
{
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
Point4 res;
const Point4 &y = *this;
res[0] = y[1]*x[2]*z[3]-y[1]*x[3]*z[2]-x[1]*y[2]*z[3]+
x[1]*y[3]*z[2]+z[1]*y[2]*x[3]-z[1]*y[3]*x[2];
res[1] = y[0]*x[3]*z[2]-z[0]*y[2]*x[3]-y[0]*x[2]*
z[3]+z[0]*y[3]*x[2]+x[0]*y[2]*z[3]-x[0]*y[3]*z[2];
res[2] = -y[0]*z[1]*x[3]+x[0]*z[1]*y[3]+y[0]*x[1]*
z[3]-x[0]*y[1]*z[3]-z[0]*x[1]*y[3]+z[0]*y[1]*x[3];
res[3] = -z[0]*y[1]*x[2]-y[0]*x[1]*z[2]+x[0]*y[1]*
z[2]+y[0]*z[1]*x[2]-x[0]*z[1]*y[2]+z[0]*x[1]*y[2];
return res;
}
//@}
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
//@{
/** @name Norms and normalizations
**/
/// Euclidian normal
inline T Norm() const
{
return math::Sqrt( _v[0]*_v[0] + _v[1]*_v[1] + _v[2]*_v[2] + _v[3]*_v[3] );
}
/// Squared euclidian normal
inline T SquaredNorm() const
{
return _v[0]*_v[0] + _v[1]*_v[1] + _v[2]*_v[2] + _v[3]*_v[3];
}
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
/// Euclidian normalization
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
inline Point4 & Normalize()
{
T n = sqrt(_v[0]*_v[0] + _v[1]*_v[1] + _v[2]*_v[2] + _v[3]*_v[3] );
if(n>0.0) { _v[0] /= n; _v[1] /= n; _v[2] /= n; _v[3] /= n; }
return *this;
}
/// Homogeneous normalization (division by W)
inline Point4 & HomoNormalize(){
if (_v[3]!=0.0) { _v[0] /= _v[3]; _v[1] /= _v[3]; _v[2] /= _v[3]; _v[3]=1.0; }
return *this;
};
//@}
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
//@{
/** @name Comparison operators (lexicographical order)
**/
inline bool operator == ( const Point4& p ) const
{
return _v[0]==p._v[0] && _v[1]==p._v[1] && _v[2]==p._v[2] && _v[3]==p._v[3];
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
}
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
inline bool operator != ( const Point4 & p ) const
{
return _v[0]!=p._v[0] || _v[1]!=p._v[1] || _v[2]!=p._v[2] || _v[3]!=p._v[3];
}
inline bool operator < ( Point4 const & p ) const
{
return (_v[3]!=p._v[3])?(_v[3]<p._v[3]):
(_v[2]!=p._v[2])?(_v[2]<p._v[2]):
(_v[1]!=p._v[1])?(_v[1]<p._v[1]):
(_v[0]<p._v[0]);
}
inline bool operator > ( const Point4 & p ) const
{
return (_v[3]!=p._v[3])?(_v[3]>p._v[3]):
(_v[2]!=p._v[2])?(_v[2]>p._v[2]):
(_v[1]!=p._v[1])?(_v[1]>p._v[1]):
(_v[0]>p._v[0]);
}
inline bool operator <= ( const Point4 & p ) const
{
return (_v[3]!=p._v[3])?(_v[3]< p._v[3]):
(_v[2]!=p._v[2])?(_v[2]< p._v[2]):
(_v[1]!=p._v[1])?(_v[1]< p._v[1]):
(_v[0]<=p._v[0]);
}
inline bool operator >= ( const Point4 & p ) const
{
return (_v[3]!=p._v[3])?(_v[3]> p._v[3]):
(_v[2]!=p._v[2])?(_v[2]> p._v[2]):
(_v[1]!=p._v[1])?(_v[1]> p._v[1]):
(_v[0]>=p._v[0]);
}
//@}
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
//@{
/** @name Dot products
**/
// dot product
inline T operator * ( const Point4 & p ) const
{
return _v[0]*p._v[0] + _v[1]*p._v[1] + _v[2]*p._v[2] + _v[3]*p._v[3];
}
inline T dot( const Point4 & p ) const { return (*this) * p; }
Rolled back
2011-10-05 17:04:40 +02:00
inline Point4 operator ^ ( const Point4& /*p*/ ) const
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
{
assert(0);// not defined by two vectors (only put for metaprogramming)
return Point4();
}
/// slower version, more stable (double precision only)
T StableDot ( const Point4<T> & p ) const
{
T k0=_v[0]*p._v[0], k1=_v[1]*p._v[1], k2=_v[2]*p._v[2], k3=_v[3]*p._v[3];
int exp0,exp1,exp2,exp3;
frexp( double(k0), &exp0 );frexp( double(k1), &exp1 );
frexp( double(k2), &exp2 );frexp( double(k3), &exp3 );
if (exp0>exp1) { math::Swap(k0,k1); math::Swap(exp0,exp1); }
if (exp2>exp3) { math::Swap(k2,k3); math::Swap(exp2,exp3); }
if (exp0>exp2) { math::Swap(k0,k2); math::Swap(exp0,exp2); }
if (exp1>exp3) { math::Swap(k1,k3); math::Swap(exp1,exp3); }
if (exp2>exp3) { math::Swap(k2,k3); math::Swap(exp2,exp3); }
return ( (k0 + k1) + k2 ) +k3;
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
}
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
//@}
}; // end class definition
template <class T>
T Angle( const Point4<T>& p1, const Point4<T> & p2 )
{
T w = p1.Norm()*p2.Norm();
if(w==0) return -1;
T t = (p1*p2)/w;
if(t>1) t=1;
return T( math::Acos(t) );
}
template <class T>
inline T Norm( const Point4<T> & p )
{
return p.Norm();
}
template <class T>
inline T SquaredNorm( const Point4<T> & p )
{
return p.SquaredNorm();
}
template <class T>
inline T Distance( const Point4<T> & p1, const Point4<T> & p2 )
{
return Norm(p1-p2);
}
template <class T>
inline T SquaredDistance( const Point4<T> & p1, const Point4<T> & p2 )
{
return SquaredNorm(p1-p2);
}
/// slower version of dot product, more stable (double precision only)
template<class T>
double StableDot ( Point4<T> const & p0, Point4<T> const & p1 )
{
return p0.StableDot(p1);
make point2 derived Eigen's Matrix, and a set of minimal fixes to make meshlab compile with both old and new version. The fixes include: - dot product: vec0 * vec1 => vec0.dot(vec1) (I added .dot() to the old Point classes too) - Transpose: Transpose is an Eigen type, so we cannot keep it if Eigen is used. Therefore I added a .tranpose() to old matrix classes, and modified most of the Transpose() to transpose() both in vcg and meshlab. In fact, transpose() are free with Eigen, it simply returns a transpose expression without copies. On the other be carefull: m = m.transpose() won't work as expected, here me must evaluate to a temporary: m = m.transpose().eval(); However, this operation in very rarely needed: you transpose at the same sime you set m, or you use m.transpose() directly. - the last issue is Normalize which both modifies *this and return a ref to it. This behavior don't make sense anymore when using expression template, e.g., in (a+b).Normalize(), the type of a+b if not a Point (or whatever Vector types), it an expression of the addition of 2 points, so we cannot modify the value of *this, since there is no value. Therefore I've already changed all those .Normalize() of expressions to the Eigen's version .normalized(). - Finally I've changed the Zero to SetZero in the old Point classes too.
2008-10-28 01:59:46 +01:00
}
Port to eigen2: state of the mess: * curently nothing change if you don't define VCG_USE_EIGEN * make Matrix*, Point3 and Point4 derive Eigen::Matrix (still ugly) * now catching all the dot products to replace them by .dot() note that most of meshlab already compile
2008-10-27 20:35:17 +01:00
typedef Point4<short> Point4s;
typedef Point4<int> Point4i;
typedef Point4<float> Point4f;
typedef Point4<double> Point4d;
/*@}*/
} // end namespace
#endif