Question regarding orthogonal complement.

inner-productslinear algebraorthogonalityprojection

We know that for any subspace $W$ of an inner product space $V$,we have $W \subset {W^\perp}^{\perp}$.We also know that for $W$ finite dimensional $W={W^\perp}^\perp$.Now I think the equality will hold whenever the orthogonal projection map of $V$ on $W$ exists ,i.e. each vector in $V$ has an orthogonal projection in $W$.Am I correct?

Best Answer

Yes, you are indeed correct. Specifically, $(W^\perp)^\perp = \operatorname{\overline{span}} W$ (the closure of the span of $W$), for any subset $W$ of $V$, not necessarily a subspace. We therefore get $(W^\perp)^\perp = W$ if and only if $W$ is a closed subspace (recall that all finite-dimensional subspaces are closed).

As for orthogonal projection onto a subspace $W$, they too exist if and only if $W$ is closed. Recall that orthogonal projection produces the closest point in $W$ to a given point $x$. If $x \in \overline{W} \setminus W$, then there cannot be a closest point in $W$ to $x$!

Related Question