Content
Summary
The row space and column space of a matrix are subspaces spanned by its rows and columns, respectively, with the rank defined as their shared dimension. While row operations preserve the row space and column dependency relations, they alter the column space, requiring careful selection of basis vectors from the original matrix. The Rank-Nullity Theorem links the rank and nullity, ensuring the sum of dimensions equals the number of columns.
Intuition
Think of a matrix as a machine that transforms input vectors into output vectors. The column space represents all possible outputs the machine can produce; if your target vector isn't in this space, the system has no solution. The row space represents the constraints or rules the machine follows. Row reduction is like simplifying the machine's instructions without changing the fundamental rules (row space) or the relationships between parts (column dependencies). However, the actual physical parts (columns) change during simplification, so you must look back at the original machine to find the independent parts that form the basis. The Rank-Nullity Theorem is a conservation law: the total number of input variables is split between those that determine the output (rank) and those that can vary without changing the output (nullity).
Key equations
The Rank-Nullity Theorem, where n is the number of columns.
Derivation
- Define row space as span of rows and column space as span of columns.
- Establish that row operations create linear combinations of rows, preserving the row space.
- Define rank as the dimension of the row space, found by counting nonzero rows in echelon form.
- Show that column dependency relations are preserved under row operations.
- Prove that row and column space dimensions are equal (rank).
- Relate nullity to free variables in reduced row echelon form.
- Combine rank and nullity to derive the Rank-Nullity Theorem.
Worked example
Given a matrix with rank 2, determine the dimension of its null space and explain how to find a basis for its column space using its row echelon form .
- Identify (number of columns) and .
- Apply the Rank-Nullity Theorem: .
- Examine the row echelon form to identify pivot columns.
- Select the corresponding columns from the original matrix (not ) to form the basis for the column space.
Answer The null space dimension is 2. The column space basis consists of the columns from that correspond to the pivot positions in .
Common mistakes
- Using the columns of the row echelon form as the basis for the column space of .
- Believing that the column space remains unchanged after row reduction.
- Confusing consistency conditions (b in column space) with uniqueness conditions (linear independence of columns).