The volume of a parallelepiped and the Gram determinant #
In a finite-dimensional real inner product space, the volume measure gives measure one to the
parallelepiped spanned by an orthonormal basis (OrthonormalBasis.volume_parallelepiped). For an
arbitrary family v of as many vectors as the dimension, the parallelepiped it spans has volume
√(det G), where G is the Gram matrix ⟪vᵢ, vⱼ⟫ of v; both sides vanish when v is linearly
dependent. For a basis b this says that the volume measure is √(det G) times the additive Haar
measure b.addHaar normalized by b.
This converts integrals against a basis-normalized Haar measure, such as the coordinate measure in which a Riemannian volume density is expressed, into integrals against the volume measure.
Main results #
TauCeti.volume_parallelepiped: the parallelepiped spanned byvhas volume√(det G).Module.Basis.volume_eq_smul_addHaar: the volume measure is√(det G) • b.addHaar.
References #
- F. R. Gantmacher, The Theory of Matrices, Vol. 1, Chelsea, 1959, Chapter IX, §5 (the Gram determinant as the squared volume of a parallelepiped).
In a finite-dimensional real inner product space, the parallelepiped spanned by a family of as many vectors as the dimension has volume the square root of the Gram determinant of the family.
The volume measure of a finite-dimensional real inner product space is the additive Haar
measure normalized by a basis b, scaled by the square root of the Gram determinant of b.