Basis $M=\{(3,1),(2,1)\}$. I solved that the dual basis \begin{equation} M^*=\{e^1,e^2\}=\{(1,-2),(-1,3)\} \end{equation}

Then I solved that matrix

\begin{equation} g=\left(\begin{array}{cc}1&-5\\-1&10\end{array}\right) \end{equation}

Now, my notes says that antisymmetrization matrix for purely covariant tensor (which is a bilinear form) is given as

$$\frac{1}{q!}\sum_{\pi \in S_q}\text{sgn}\pi.(e_{\pi(1)},.....,e_{\pi(n)})$$

Here $q$ is 2 (because bilinear form is a (0,2) tensor and since we're in 2 dimensional space, there should be two permutations).

If I understant correctly, my matrix should be equal to


I just don't know what these $e_1$ and $e_2$ mean (I suppose they are not meant to be the same as the basis vectors of $M$).

Are $e_1$ and $e_2$ meant to be columns of the matrix $g$?

In that case the matrix of antisymmetrization would be



Actually, the $e_{\pi(i)}$ are inteded to be the same you computed in the dual basis, but with another meaning. In fact, $(e_2,e_1)$ and $(e_1,e_2)$ are to be intended as a permutation operators on the tensor.

In particular, $(e_1,e_2)$ is the identity operator, whereas $(e_2,e_1)$ swaps the indices as $$ (e_2,e_1) (e_1\otimes e_1) =e_2 \otimes e_2,\\ (e_2,e_1) (e_1\otimes e_2) =e_2 \otimes e_1,\\\dots $$ Notice that these operators are linear, and you can compute their effect on the matrix. In fact, $(e_1,e_2)$ does not change the matrix $g$, whereas $(e_2,e_1)$ coincides with the transposition of the matrix.

In your case, $$ \pi_A(g)=\frac{1}{2}(g-g^T) $$ that is what is commonly known as the skew(or anti)-symmetric part of $g$.

| cite | improve this answer | |

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.