Skip to lesson

Math · Linear Algebra

Chapter 5: Determinants

The Invertibility Theorem

Eight statements that are all the same statement.

Lesson
3
Time
About 24 minutes
0 of 12 done

Step 1: Let's Learn

Read it, or press Listen and follow the words.

Several apparently different questions about a square matrix always have the same answer. Learning them as one fact saves enormous effort.

The invertible side

A has an inverse. Its determinant is non-zero. Its columns are independent. Its columns span the whole space.

And also

Ax = b has exactly one solution for every b. Ax = 0 has only the zero solution. Every column has a pivot.

The other side

If any one of those fails, all of them fail. There is no matrix with a zero determinant that still has independent columns.

Why they agree

Each statement describes the same event: whether the transformation flattened space or not. Flattening loses information permanently.

How to use it

Check whichever condition is cheapest. For a small matrix the determinant is usually the fastest door in.

Many statements, one condition

For a square matrix, being invertible, having nonzero determinant, having independent columns, having full rank, and Ax = b always being solvable are all the same condition in different language.

Why the theorem is useful

Whichever version is easiest to check settles all the others. Establishing independence proves solvability without solving anything, which is a genuine economy.

It applies to square matrices only

A non-square matrix has no inverse and no determinant. The equivalences are a theorem about square matrices specifically, and applying them elsewhere is a category error.

Nearly singular is a practical problem

A determinant close to zero means the matrix is almost non-invertible, and solutions become extremely sensitive to small errors. In computation, near-singularity matters more than exact singularity.

Step 2: Try It Yourself

Tap and try it out.

Every claim in this lesson is about this one picture. When the parallelogram survives, all eight statements hold; when it collapses, all eight fail.
ij
  • i-hat lands on(2, 4)
  • j-hat lands on(1, 2)
  • Determinant0

The parallelogram has collapsed to a segment. The determinant is zero, the whole plane has been squashed onto a line, and no inverse can undo it.

Step 3: Watch an Example

One step at a time.

Watch Omar Settle Four Questions at Once

Omar is asked whether a matrix with rows [2, 4] and [1, 2] is invertible, has independent columns, spans the plane, and gives unique solutions.

  1. Step 1

    He computes the determinant: 4 − 4 = 0.

Step 4: Your Turn

Practice makes it stick.

The One Check

Problem 1 of 2

A square matrix has determinant 0. How many solutions does Ax = 0 have beyond the zero vector? Enter 0 for none, 1 for infinitely many.

The Rank

Problem 2 of 2

An invertible 4 by 4 matrix. How many pivots does it have?

One Fact, Many Faces

1 of 8

Determinant 5. Is the matrix invertible? 1 yes, 0 no.

2 of 8

Determinant 0. Are the columns independent? 1 yes, 0 no.

3 of 8

An invertible 3 by 3 matrix. How many pivots?

4 of 8

Ax = 0 has only the zero solution. Is A invertible? 1 yes, 0 no.

5 of 8

Rows [3, 6] and [1, 2]. Is the matrix invertible? 1 yes, 0 no.

6 of 8

The columns span the whole space. Is the determinant non-zero? 1 yes, 0 no.

7 of 8

Sort each statement by which side of the theorem it belongs to.

Tap something to move it.

  • Empty
  • Empty

8 of 8

Rows [1, 0] and [0, 1]. Is the matrix invertible? 1 yes, 0 no.

Step 5: Quick Check

Show what you know.

Question 1 of 2

Rows [4, 8] and [1, 2]. Is the matrix invertible? 1 yes, 0 no.

Question 2 of 2

Why do all these conditions agree?

What You Learned

  • For a square matrix, invertibility, a non-zero determinant, independent columns and unique solutions are all one fact.
  • If one condition fails, every one of them fails.
  • They agree because they all describe whether the transformation flattened space.