|
Developer Documentation |
|
Centre for Vision, Speech & Signal Processing |
| LinearSums2dC
|
| |
| The class LinearSums2dC serves for computation of linear dependences in 2D space.
|
| |
| include | "amma/LinearS2.hh" |
| User Level: | Default |
| Library: | Mag2 |
| Example: | exMat2d.cc |
| Section: | Geometry.2-D |
| In Scope: | std |
Typedefs:
- typedef StdTypeC::RealT RealT;
-
Variables:
- LongIntT n;
-
- RealT sumX;
-
- RealT sumY;
-
- RealT sumXX;
-
- RealT sumXY;
-
- RealT sumYY;
-
Methods:
- LinearSums2dC()
-
The zero sums.
- LinearSums2dC(const LinearSums2dC & sums)
-
Copy constructor.
Updating.
---------
- void SetZero()
-
Initializies the sums.
- const LinearSums2dC & operator+=(const Point2dC & p)
-
Adds another point to the sums.
- const LinearSums2dC & operator-=(const Point2dC & p)
-
Subtracts the point from the sums.
- const LinearSums2dC & Add(const RealT x,const RealT y)
-
Adds another point to the sums.
- const LinearSums2dC & Remove(const RealT x,const RealT y)
-
Removes the point from the sums.
Computation of statistics.
--------------------------
- LongIntT N() const
-
Returns the number of data have been entered into the sums.
- RealT MeanX() const
-
Returns the mean value of the x.
- RealT MeanY() const
-
Returns the mean value of the y.
- RealT VarX() const
-
Returns the variance of the x.
- RealT VarY() const
-
Returns the variance of the y.
- Point2dC Centroid() const
-
Returns the point with the mininum distance from this set of points.
The used criterion is Sum((Y-y)^2+(X-x)^2) -> min.
- RealT SlopeY() const
-
Returns the slope dY/dX. The used criterion is Sum(Y-y)^2 -> min.
It means dY/dX = k, where y = k*x+q.
- RealT SlopeX() const
-
Returns the slope dX/dY. The used criterion is Sum(X-x)^2 -> min.
It means dX/dY = k, where x = k*y+q.
- RealT InterceptY() const
-
Returns the estimate of q, if y = k*x+q.
The used criterion is Sum(Y-y)^2 -> min.
- RealT InterceptX() const
-
Returns the estimate of q, if y = k*y+q.
The used criterion is Sum(X-x)^2 -> min.
- RealT Angle() const
-
If points form a line, estimate its angle.
Returns an angle in randians.
- LineABC2dC LineABCy() const
-
Returns the equation of the line. The used criterion is
Sum(Y-y)^2 -> min.
- LineABC2dC LineABCx() const
-
Returns the equation of the line. The used criterion is
Sum(X-x)^2 -> min.
- BooleanT UseLineABCx() const
-
Which is better LineABCx() or LineABCy ?
- LineABC2dC LineABCxy() const
-
Return the best fit of LineABCx, LineABCy.
This may change to Sum(Y-y)^2 + Sum(X-x)^2 -> min
in the future.
- RealT MinimumVolume() const
-
Returns the determinant of the centralized covariance matrix.
The determinant is propotional to the volume of the minimum
elipsoid containing data.
- RealT Correlation() const
-
Returns the value of the correlation between X and Y.
- ostream & operator<<(ostream & outS,const LinearSums2dC & sums)
-
- istream & operator>>(istream & inS,LinearSums2dC & sums)
-
- const LinearSums2dC & Add(const LinearSums2dC::RealT x,const LinearSums2dC::RealT y)
-
- const LinearSums2dC & Remove(const LinearSums2dC::RealT x,const LinearSums2dC::RealT y)
-
|
Programmer:Radek Marik, Documentation by CxxDoc: Tue Mar 20 10:49:27 2001
|