Fix #522: float-guard the identity matrix test

Provide floating-point error margins in testing whether or not a matrix
is the identity matrix. This is done on a component by component basis.
Previously, matrix components were compared with exactly '0' and '1', so
the test could fail on account of roundoff errors on the components.

HEN_HOUSE/egs++/egs_transformations.h

@@ -215,11 +215,23 @@ class EGS_EXPORT EGS_RotationMatrix {
  (rzx == m.rxz) && (rzy == m.rzy) && (rzz == m.rzz)) ? true : false;
  };
 
- /*! \brief Returns \c true, if this object is a unity matrix, \a false
+ /*! \brief Returns \c true, if this object is approximately the unit matrix, \a false
  otherwise */
  bool isI() const {
- return (rxx == 1 && rxy == 0 && rxz == 0 && ryx == 0 && ryy == 1 &&
- ryz == 0 && rzx == 0 && rzy == 0 && rzz == 1);
+
+ // compare with unit matrix components
+ if (fabs(rxy) < epsilon &&
+ fabs(rxz) < epsilon &&
+ fabs(ryx) < epsilon &&
+ fabs(ryz) < epsilon &&
+ fabs(rzx) < epsilon &&
+ fabs(rzy) < epsilon &&
+ fabs(rxx-1) < epsilon &&
+ fabs(ryy-1) < epsilon &&
+ fabs(rzz-1) < epsilon ) {
+ return true;
+ }
+ return false;
  };
 
  /*! \brief Returns the rotated a vector \f$ R \cdot \vec{v}\f$ */