← Linear Algebra

Unit 4: Vector Spaces and the Four Fundamental Subspaces

Every matrix carries four hidden spaces around with it — the column space, null space, row space, and left null space.

15 topics · 58 questions

Each topic is a path, not a menu: work down it in order and every question builds on the one above. Click any of them to watch the answer draw itself.

What a Vector Space Is

  1. How it worksHow do the rules for adding and scaling decide whether a collection of objects forms a vector space?
  2. Common mistakeWhy isn't a vector space required to be made of arrows or lists of numbers?
  3. Going furtherWhat happens to the vector space of all polynomials once you only allow degree 2 or less?
Your turn1 of 2
Matrices, polynomials, and geometric arrows all count as vector spaces. What makes them all the exact same KIND of mathematical object?

Subspaces of a Vector Space

  1. How it worksHow do closure under addition and closure under scaling decide whether a subset is actually a subspace?
  2. Worked exampleShow that the set of vectors (x, y) with y equal to 2x forms a subspace of the plane.
  3. Going furtherWhat happens when you take the union of two different lines through the origin — does it stay a subspace?
  4. Common mistakeWhy isn't a line that misses the origin ever a subspace?
Your turn1 of 2
The set of points (x, y) with y = 2x + 4, where 4 is nonzero, does not pass through the origin. Why can this never be a subspace of the plane, regardless of what k and c are?

The Column Space of a Matrix

  1. How it worksHow does the column space collect every vector you can reach by combining the columns of A?
  2. Worked exampleFind the column space of the matrix [[1, 2], [2, 4]].
  3. Common mistakeWhy isn't the column space of a matrix the same thing as the plain set of its columns?
  4. Going furtherWhat happens to the column space of a matrix when you add an extra column that's a combination of the others?
Your turn1 of 2
Why isn't the column space of a matrix the same thing as the plain set of its columns?

The Null Space of a Matrix

  1. How it worksHow do free variables in the reduced matrix generate every single vector in the null space?
  2. Worked exampleFind the null space of the matrix [[1, 2, 3], [2, 4, 6]].
  3. Worked exampleFind every vector that the matrix [[1, 1, 2], [2, 2, 4], [1, 0, 1]] sends to zero.
  4. Common mistakeWhy isn't the null space just the set of zero entries sitting inside a matrix?
  5. Going furtherWhat happens to the null space of a matrix once it has full column rank?
Your turn1 of 2
A matrix has first row (1, 3, 2) and its second row is exactly double the first row (so its rank is 1, with 2 free variables among x, y, z in its null space). Find the value of x in the null-space vector where y = 1 and z = 0.

Solving Ax = 0

  1. How it worksHow does setting each free variable to 1 in turn generate the special solutions that span the null space?
  2. Worked exampleFind the special solutions spanning the null space of the matrix [[1, 2, 0, 3], [0, 0, 1, 4]].
  3. Ties togetherHow does the null space of a matrix connect directly to whether its columns are linearly independent?
Your turn1 of 2
A matrix reduces to [[1, 3, 0, 2], [0, 0, 1, 7]]. Its null space is spanned by two special solutions, one for each free variable (x2 and x4). Find the x1 value of the special solution with x2 = 0 and x4 = 1.

The Complete Solution to Ax = b

  1. Common mistakeWhy doesn't a system with a nonzero null space mean there is no solution to Ax = b?
  2. How it worksHow does adding any vector from the null space to one particular solution generate every other solution?
  3. Worked exampleFind the complete solution to the system x + 2y = 3, 2x + 4y = 6.
  4. Going furtherWhat happens to the complete solution once the null space of A contains only the zero vector?
Your turn1 of 2
The system x + 2y = 3, with its second equation being a multiple of the first (so infinitely many solutions lie along one line), has null-space direction (-2, 1) and particular solution (3, 0). Find the x-coordinate of the general solution at parameter t = 5.

Linear Independence

  1. How it worksHow does forcing c1v1 + c2v2 + ... + cnvn = 0 to have only the all-zero solution test for independence?
  2. Worked exampleShow that the vectors (1, 0, 1), (0, 1, 1), and (1, 1, 0) are linearly independent.
  3. Common mistakeWhy doesn't checking that no two vectors are multiples of each other prove three or more vectors are independent?
  4. Going furtherWhat happens to independence the moment a fourth vector joins three vectors that already span all of 3-space?
Your turn1 of 2
The vectors (1, 0, 2), (0, 1, 2), and (1, 5, 0) are independent exactly when the determinant of the matrix they form is nonzero. Find that determinant.

A Basis of a Vector Space

  1. How it worksHow does a basis have to be both independent and span the whole space at the same time?
  2. Worked exampleShow that (1, 0) and (1, 1) form a basis for the plane.
  3. Common mistakeWhy doesn't a spanning set automatically count as a basis?
  4. Going furtherWhat happens when you try to build a basis for the plane out of three vectors instead of two?
Your turn1 of 2
Find ad - bc for the vectors (4, 3) and (8, -4), which form a basis of the plane exactly when this value is nonzero.

The Dimension of a Vector Space

  1. How it worksHow does counting the vectors in any one basis give you the dimension of the whole space?
  2. Worked exampleFind the dimension of the subspace of the plane spanned by (2, 4) and (1, 2).
  3. Common mistakeWhy isn't the dimension of a vector space just how many numbers it takes to write one vector down?
  4. Ties togetherHow does the dimension of the column space connect back to the rank of a matrix?
Your turn1 of 2
Find the dimension of the column space of a matrix that has 6 pivot columns after row reduction.

Row Rank Equals Column Rank

  1. How it worksHow does elimination preserve the row space while collapsing dependent columns, proving the two ranks must agree?
  2. Worked exampleShow that row rank equals column rank for the matrix [[1, 2, 3], [2, 4, 7]].
  3. Common mistakeWhy isn't the row space of a matrix literally equal to its column space, even though their dimensions always match?
  4. Going furtherWhat happens to this row-rank-equals-column-rank equality when the matrix is rectangular instead of square?
Your turn1 of 2
Show that row rank equals column rank for the matrix [[1, 3, 2], [0, 1, 7]]. A general vector in the row space is s(1, a, b) + t(0, 1, c) for scalars s and t. Find its second entry when s = -3 and t = -5.

The Row Space and Left Null Space

  1. How it worksHow does the row space collect every combination of the rows of A, the same way the column space combines columns?
  2. Worked exampleFind the row space and left null space of the matrix [[1, 2], [2, 4]].
  3. Common mistakeWhy isn't the row space of a matrix the same as its column space, even when they share the same dimension?
  4. Going furtherWhat happens to the left null space of a matrix that has full row rank?
Your turn1 of 2
A 2 by 2 matrix has first row (3, 2) and its second row is exactly 6 times the first row. Find the second entry of a left-null-space vector whose first entry is 1.

The Rank of a Matrix

  1. How it worksHow does the number of pivots left after elimination tell you the rank of a matrix?
  2. Worked exampleFind the rank of the matrix [[1, 2, 3], [2, 4, 6], [1, 1, 1]].
  3. Common mistakeWhy isn't the rank of a matrix just the number of rows it has?
  4. Going furtherWhat happens to the rank of a matrix when you add a row that's a combination of the existing rows?
Your turn1 of 2
A matrix currently has rank 5. A new row is added that is NOT a linear combination of the existing rows. Find the new rank.

The Rank-Nullity Theorem

  1. How it worksHow does the number of pivot columns plus the number of free columns always add up to the total column count?
  2. Worked exampleShow the rank-nullity theorem holds for the matrix [[1, 2, 3, 4], [2, 4, 6, 9]].
  3. Common mistakeWhy doesn't a bigger null space mean a matrix has to have a smaller column space?
Your turn1 of 2
A matrix has 7 columns and rank 2. Find its nullity — the dimension of its null space.

Existence and Uniqueness via Rank

  1. How it worksHow does comparing rank(A) to the rank of the augmented matrix [A given b] tell you whether a solution exists at all?
  2. Worked exampleFind the number of solutions of the system x + y = 2, 2x + 2y = 5 by comparing the rank of A to the rank of the augmented matrix.
  3. Common mistakeWhy doesn't a system with more equations than unknowns automatically mean there is no solution?
Your turn1 of 2
A system has 7 unknowns. Its coefficient matrix A has rank 2, and the augmented matrix [A | b] also has rank 2 (so the system is consistent). Find the number of free variables in the solution.

The Four Fundamental Subspaces

  1. How it worksHow do the column space, null space, row space, and left null space each live in either the input space or the output space?
  2. Worked exampleFind the dimensions of all four fundamental subspaces of the matrix [[1, 2, 3], [2, 4, 6]].
  3. Common mistakeWhy doesn't the row space of A end up equal to the column space of A, even though the two have the same dimension?
  4. Going furtherWhat happens to the four fundamental subspaces when the matrix A is square and invertible?
  5. Ties togetherHow does orthogonality between the four subspaces set up the picture used later for projections?
Your turn1 of 2
A matrix is 7 by 7 with rank 3. Find the dimension of its left null space.