I've read some book that use the wedge products of $dx_i$ to define $r$-forms in $\mathbb{R}^n$. However, I have a professor that defines the $r$-forms just using that are $r$-linear and antisymmetric.
How can I get from here that the wedge products of lenght $r$ generate the differential $r$-forms?
Many thanks in advance.
This is a fundamental fact from multilinear algebra. I'll follow Tu's "Introduction to Manifolds," and present it from a more general perspective.
First, say we have a basis $e_1,\cdots e_n$ for a real vector space $V$, and let $\alpha^1,\cdots,\alpha^n$ be the dual basis for $V'.$ First, one proves a lemma
The proof is not hard, and it uses the relationship between evaluating a wedge of $k$ covectors and the determinant. Using such a lemma, one can proceed to show that
The lemma is used to prove linear independence. Showing that they span is straightforward. For details, I recommend reading Tu (pages $31$-$32$).
In your situation, $V=T_p(\mathbb{R}^n),$ and $e_j=\partial/\partial_{x_j}|_p,$ $\alpha^j=dx^j|_p.$