LightMySky
En anglais

Change of Basis

The same map has different matrices in different bases. A change-of-basis matrix converts coordinates, and similar matrices are the same map seen from two viewpoints.

No account needed. Progress saves in this browser.

What a learner can do afterwards

  • Convert coordinates between two bases with a transition matrix
  • Write the matrix of a map in a new basis by conjugating
  • Explain what stays the same about a map when the basis changes

1 · Read

The same vector has different coordinates in different bases, and the transition matrix is the dictionary between the languages. Its columns are the old basis vectors written in the new coordinates. Example: (5, 1) = 3(1, 1) + 2(1, minus 1), so its B-coordinates are (3, 2). Label every vector with its basis.

Try it together

The same map has different matrices too: A new = P inverse A P, read as translate, apply, translate back. Here P has the new basis vectors as its columns, so P turns new coordinates into old ones. With A = [[1, 2], [0, 3]] and P = [[1, 1], [0, 1]], the new matrix is [[1, 0], [0, 3]], whose (1, 2) entry is 0. Conversion runs through an inverse, so it works in both directions.

Try it together

To fit a projection, use a basis along its line and across it. Projection onto the line y = x keeps (1, 1) and sends (1, minus 1) to zero. In the basis (1, 1), (1, minus 1), those images have new coordinates (1, 0) and (0, 0), so the matrix is [[1, 0], [0, 0]]: keep the first coordinate, drop the second. In standard coordinates the same map is [[0.5, 0.5], [0.5, 0.5]].

(1, 1) goes to (1, 1)(1, -1) goes to (0, 0)images: (1, 0), (0, 0)matrix [[1, 0], [0, 0]]
Each basis vector's image, written in the new basis, becomes one column.

The map stays fixed while only its description changes, so some quantities are immune. The trace, the sum of the diagonal entries, and the determinant belong to the transformation itself. So do eigenvalues: a number lambda is an eigenvalue when A v = lambda v for some nonzero v, which describes what the map does, not how the matrix is written. The diagonal matrix [[3, 0], [0, 8]] has trace 3 + 8 = 11 in every basis, and no basis change moves its determinant, 24.

Good to know

Most change of basis errors come from untracked labels. Write the basis next to every coordinate vector before you multiply. If a result looks strange, check first whether two languages got mixed in one equation.

Convert coordinates with P, move matrices with P inverse A P, and trust trace, determinant, and eigenvalues to survive the trip.

Reading notes describe open textbooks listed on OER Commons. Videos by Khan Academy. Practice questions are original.

2 · Watch

En anglais

Take it off screen

Print a worksheetA4 with an answer key page for grown-ups. No screen, no internet.

Where it sits

Then practise

14 questions wait behind this lesson, each with its answer explained. Every answer feeds the sky: stars light as they are learned, and dim when it is time to come back.

Spotted a problem on this page? Tell us
Change of Basis · Mathematics, ages 19 to 20 · LightMySky