Apparently the question was answered in the comments: Edmonds proves a more general theorem for a family of matroids. If the matroids are all the same and if that matroid is a vector matroid, then it is exactly the question stated.
Greg Kuperberg
- 56.6k
- 10
- 203
- 282