Hodge dual computations#

by Stéphane Haussler

Warning

Draft

In this page, I present a straightforward algorithmic method to calculate the inner product between k–vectors (or k–forms). The dicussion in the article on Hodge duality fundamentals provides geometric intuition. Here, I offer an alternative approach to compute the inner product of k–forms, as well as the Hodge dual of k–forms, with a simple algorithmic method using the interior product \(⌟\).

My use of the interior product might be non-standard, and I have not investigated whether what follows is established. It certainly works as I intuitively expect it should. You may also just find this method a usefull as an efficient trick to determine the inner product between k–forms, or the Hodge dual of k-forms.

In this page, I call the top-dimensional differential form the pseudo-scalar. Specifically, the pseudo-scalar is:

  • \(dx ∧ dy ∧ dz\) in 3D–Euclidean space, and

  • \(dt ∧ dx ∧ dy ∧ dz\) in 4D–Minkowski space.

The term pseudo-scalar is chosen because the Hodge dual of these top-dimensional differential forms results in scalar quantities. This naming convention highlights the relationship between these forms and scalar fields through the Hodge duality.

Inner product on k–froms#

To establish a common foundation, recall that the inner product of k–vectors is equal to that of k–forms. For basis vectors and covectors, the inner product is the metric. In flat spacetime, this yields the Minkowski metric \(η\):

\[∂_μ · ∂_ν = dx^μ · dx^ν = η^{μν} = η_{μν}\]

Considering basis vectors, dual covectors are defined as:

\[dx^μ \left( ∂_ν \right) = δ^μ_ν\]

In this sense, 1–forms measure vectors, and the measure of a basis vector \(∂_ν\) through it corresponding basis covector \(dx^μ\) is one. The inner product, denoted with \(·\) or \(\braket{|}\), measures the shadow of one vector onto another.

The standard interior product is an operation between a vector and a form. It consist of rearanging the form to bring the corresponding covector to the front, and applying the vector to that front slot. This operation transforms a k–form to a (k-1)–form. For example:

\[\newcommand{\⌟}{\:⌟\:} ∂_y \⌟ dx ∧ dy ∧ dz = ∂_y \⌟ (- dy ∧ dx ∧ dz) = - dy \left( ∂_y \right) ∧ dx ∧ dz = - dx ∧ dz\]

Here, we have applied the basis vector \(∂_ν\) to the basis form \(dx^μ\).

We equate the interior product to the dot product for a vector pair to the interior product \(⌟\):

\[\newcommand{\⌟}{\:⌟\:} ∂_ν \⌟ ∂_μ = ∂_μ · ∂_ν = η_{μν}\]

For covectors:

\[\begin{split}\newcommand{\⌟}{\:⌟\:} dx^ν \⌟ dx^μ = dx^μ · dx^ν = η^{μν} \\\end{split}\]

The interior product of a vector can be applied to a general k–vector. For a basis vector acting on a basis 2–vector, we bring the corresponding basis 1–vector to the front and apply the vector to the first slot:

\[\begin{split}\newcommand{\⌟}{\:⌟\:} dt \⌟ dx ∧ dt &= dt \⌟ \left( - dt ∧ dx \right) \\ &= - ( dt · dt ) \; dx \\ &= - dx\end{split}\]

Simmilarly, the interior product of a basis vector can be taken with a 3–vector or a 4–vector. This provides a straightforward algorithmic approach to calculate the inner product with no thinking involved:

\[\begin{split}\newcommand{\⌟}{\:⌟\:} (dt ∧ dx) · (dt ∧ dx) &= dx \⌟ dt \⌟ dt ∧ dx \\ &= dx \⌟ (dt · dt) ∧ dx \\ &= dx \⌟ dx \\ &= dx · dx \\ &= -1\end{split}\]

Hodge dual of k–forms#

In the article Hodge duality fundamentals, the inclusion of the inner product to determine the Hodge dual of k–forms is explained. Here I present how to directly determine the Hodge dual. The Hodge dual can be determined directly by taking the interior product \(⌟\) with the pseudo-scalar. In Minkowski space, the pseudo-scalar is \(dt ∧ dx ∧ dy ∧ dz\).

For example, the Hodge dual of \(dx ∧ dy\) is:

\[\begin{split}\newcommand{\⌟}{\:⌟\:} ⋆ dx ∧ dy & = (dx ∧ dy) \⌟ dt ∧ dx ∧ dy ∧ dz \\ & = dy \⌟ dx \⌟ (- dx ∧ dt ∧ dy ∧ dz) \\ & = dy \⌟ dx \⌟ (+ dx ∧ dy ∧ dt ∧ dz) \\ & = dy \⌟ (dx · dx) \: dy ∧ dt ∧ dz \\ & = - dy \⌟ dy ∧ dt ∧ dz \\ & = - (dy · dy) \: dt ∧ dz \\ & = dt ∧ dz \\\end{split}\]

Euclidean space#

Inner product#

Hodge duals#

Minkowski space#

Inner product#

We can systematicall apply the procedure to obtain the same result as above:

1–forms

\[\begin{split}\begin{array}{c|rrrr} & dt & dx & dy & dz \\ \hline dt & +1 & 0 & 0 & 0 \\ dx & 0 & -1 & 0 & 0 \\ dy & 0 & 0 & -1 & 0 \\ dz & 0 & 0 & 0 & -1 \\ \end{array}\end{split}\]
\[\begin{split}\newcommand{\⌟}{\:⌟\:} \begin{alignedat}{5} dt · dt &=& dt &\⌟& dt & = +1 \\ dx · dt &=& dx &\⌟& dx & = -1 \\ dy · dt &=& dy &\⌟& dy & = -1 \\ dz · dt &=& dz &\⌟& dz & = -1 \\ \end{alignedat}\end{split}\]

2–forms

\[\begin{split}\begin{array}{c|rrrrrr} & dt ∧ dx & dt ∧ dy & dt ∧ dz & dy ∧ dz & dz ∧ dx & dx ∧ dy \\ \hline dt ∧ dx & -1 & 0 & 0 & 0 & 0 & 0 \\ dt ∧ dy & 0 & -1 & 0 & 0 & 0 & 0 \\ dt ∧ dz & 0 & 0 & -1 & 0 & 0 & 0 \\ dy ∧ dz & 0 & 0 & 0 & +1 & 0 & 0 \\ dz ∧ dx & 0 & 0 & 0 & 0 & +1 & 0 \\ dx ∧ dy & 0 & 0 & 0 & 0 & 0 & +1 \\ \end{array}\end{split}\]
\[\begin{split}\newcommand{\⌟}{\:⌟\:} \newcommand{\·}{\:·\:} \begin{alignedat}{5} (& dt ∧ dx &) \· (& dt ∧ dx &) =& dx &\⌟& dt &\⌟& dt ∧ dx &= + dx &\⌟& dx &= -1 \\ (& dt ∧ dy &) \· (& dt ∧ dy &) =& dy &\⌟& dt &\⌟& dt ∧ dy &= + dy &\⌟& dy &= -1 \\ (& dt ∧ dy &) \· (& dt ∧ dz &) =& dz &\⌟& dt &\⌟& dt ∧ dz &= + dz &\⌟& dz &= -1 \\ (& dy ∧ dz &) \· (& dy ∧ dz &) =& dz &\⌟& dy &\⌟& dy ∧ dz &= - dz &\⌟& dz &= +1 \\ (& dz ∧ dx &) \· (& dz ∧ dx &) =& dx &\⌟& dz &\⌟& dz ∧ dx &= - dx &\⌟& dx &= +1 \\ (& dx ∧ dy &) \· (& dx ∧ dy &) =& dy &\⌟& dx &\⌟& dx ∧ dy &= - dy &\⌟& dy &= +1 \\ \end{alignedat}\end{split}\]

3–forms

\[\begin{split}\begin{array}{c|rrrr} & dx ∧ dy ∧ dz & dt ∧ dy ∧ dz & dt ∧ dz ∧ dx & dt ∧ dx ∧ dy \\ \hline dx ∧ dy ∧ dz & -1 & 0 & 0 & 0 \\ dt ∧ dy ∧ dz & 0 & +1 & 0 & 0 \\ dt ∧ dz ∧ dx & 0 & 0 & +1 & 0 \\ dt ∧ dx ∧ dy & 0 & 0 & 0 & +1 \\ \end{array}\end{split}\]
\[\begin{split}\newcommand{\⌟}{\:⌟\:} \newcommand{\·}{\:·\:} \small \begin{alignedat}{5} (& dx ∧ dy ∧ dz &) \· (& dx ∧ dy ∧ dz &) =& dz \⌟ dy \⌟ dx \⌟ dx ∧ dy ∧ dz &=& - dz \⌟ dy \⌟ dy ∧ dz &= + dz \⌟ d = -1 \\ (& dt ∧ dy ∧ dz &) \· (& dt ∧ dy ∧ dz &) =& dz \⌟ dy \⌟ dt \⌟ dt ∧ dy ∧ dz &=& + dz \⌟ dy \⌟ dy ∧ dz &= - dz \⌟ d = +1 \\ (& dt ∧ dz ∧ dx &) \· (& dt ∧ dz ∧ dx &) =& dx \⌟ dz \⌟ dt \⌟ dt ∧ dz ∧ dx &=& + dx \⌟ dz \⌟ dz ∧ dx &= - dx \⌟ d = +1 \\ (& dt ∧ dx ∧ dy &) \· (& dt ∧ dx ∧ dy &) =& dy \⌟ dx \⌟ dt \⌟ dt ∧ dx ∧ dy &=& + dy \⌟ dx \⌟ dx ∧ dy &= - dy \⌟ d = +1 \\ \end{alignedat}\end{split}\]

4–forms

\[\begin{split}\begin{array}{c|c} & dt ∧ dx ∧ dy ∧ dz \\ \hline dt ∧ dx ∧ dy ∧ dz & -1 \\ \end{array}\end{split}\]
\[\begin{split}\newcommand{\⌟}{\:⌟\:} \newcommand{\·}{\:·\:} (dt ∧ dx ∧ dy ∧ dz) \· (dt ∧ dx ∧ dy ∧ dz) &= dz \⌟ dy \⌟ dx \⌟ dt \⌟ dt ∧ dx ∧ dy ∧ dz \\ &= dz \⌟ dy \⌟ dx \⌟ dx ∧ dy ∧ dz \\ &= -1 dz \⌟ dy \⌟ ∧ dy ∧ dz \\ &= +1 dz \⌟ ∧ dz \\ &= -1\end{split}\]

Hodge duals#

1-forms

\[\begin{split}⋆ dt &=& dx ∧ dy ∧ dz \\ ⋆ dx &=& dt ∧ dy ∧ dz \\ ⋆ dy &=& dt ∧ dz ∧ dx \\ ⋆ dz &=& dt ∧ dx ∧ dy \\\end{split}\]

2-forms

\[\begin{split}⋆ dt ∧ dx &= \\ ⋆ dt ∧ dy &= \\ ⋆ dt ∧ dz &= \\ ⋆ dy ∧ dz &= \\ ⋆ dz ∧ dx &= \\ ⋆ dx ∧ dy &= \\\end{split}\]

3-forms

\[\begin{split}⋆ dx ∧ dy ∧ dz &= \\ ⋆ dt ∧ dy ∧ dz &= \\ ⋆ dt ∧ dz ∧ dx &= \\ ⋆ dt ∧ dx ∧ dy &= \\\end{split}\]

4-forms

\[⋆ dt ∧ dx ∧ dy ∧ dz =\]