Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. Join them; it only takes a minute:

Sign up
Here's how it works:
  1. Anybody can ask a question
  2. Anybody can answer
  3. The best answers are voted up and rise to the top

I am reading "Introduction to smooth manifolds" by Lee and one place is very unclear for me: Let $P$ and $Q$ be any complementary subspaces of $V$ (which is an $n$-dimensional real vector space) of dimensions $k$ and $n-k$, respectively, so that $V$ decomposes as a direct sum $V=P\bigoplus Q$.(This part is clear for me) The graph of any linear map $X:P\rightarrow Q$ can be identified with a $k$-dimensional subspace $\Gamma(X)\subseteq V$, defined by $$\Gamma(X)=\left \{ v+Xv:v\in P \right \}$$

Earlier in the same book there was a example of a function graph being a manifold. And the graph of the continuous function $f:U \subseteq R^{n} \rightarrow R^{k}$ is the subset of $R^{n}\times R^{k}$ defined by $$\Gamma(X)=\left \{ (x,y)\in R^{n}\times R^{k}:x\in U, y=f(x) \right \}.$$

By this definition the graph of the linear map $X$ would be a subset of $R^{k}\times R^{n-k}$, but the book says it is a $k$ dimensional subspace of $V$. And this is where I am confused. What am I overlooking?

share|cite|improve this question
up vote 1 down vote accepted

For any $v\in P$, $(v,Xv)=0$ iff $v=0$. Therefore the nullity of the linear map $f:v\mapsto(v,Xv)$ is zero and by the rank-nullity theorem, $\operatorname{rank}f\equiv\dim\{(v,Xv):v\in P\}=\dim P=k$. Hence the graph of $X$, i.e. $\{(v,Xv):v\in P\}\subset\mathbb{R}^k\times\mathbb{R}^{n-k}$, can be identified with the $k$-dimensional subspace $\{v+Xv:v\in P\}\subset\mathbb{R}^n$.

share|cite|improve this answer
This is also a very nice answer. Thanks! – Tomas Jan 18 '13 at 7:19

Since $V$ is finite dimensional you have $V= P \bigoplus Q \simeq P \times Q = R^k \times R^{n-k}$. So it is a subset of $R^k \times R^{n-k}$ and it is a $k$-vector space since

$$P \longleftrightarrow \Gamma(X) = \lbrace v+Xv : v \in P \rbrace: v \longleftrightarrow (v,X(v))$$

is an isomorphism of vector spaces.

share|cite|improve this answer
Yes, I know that, but this is not my question. My question is why a graph of linear map $X$ can be identified with $k$-dimensional subspace $\Gamma (X)$? – Tomas Jan 17 '13 at 20:50
What I try to point out is that both definitions are the same, with different notation: $v+Xv =(v, X(v))$ – Z. L. Jan 17 '13 at 20:54
Observe that a linear map is also a continuous function with respect to the topology induced by a norm in that vector space. – Z. L. Jan 17 '13 at 20:55
Since it is parametrized from $v \in P$ you can identify it with $P$ and $P$ is a k-dimensional subspace of $V$. – Z. L. Jan 17 '13 at 20:57
Thanks, now it is clear!! – Tomas Jan 18 '13 at 7:16

Your Answer


By posting your answer, you agree to the privacy policy and terms of service.

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