Change of basis matrices rewrite a vector’s coordinates from one basis to another in Linear Algebra and Differential Equations. They let you describe the same vector with a different coordinate system without changing the vector itself.
A change of basis matrix is the matrix you use to translate coordinate vectors from one basis to another in Linear Algebra and Differential Equations. The vector stays the same, but its ordered list of coordinates changes because the basis vectors you are using to measure it have changed.
Here is the core idea: a basis is a chosen set of vectors that spans the space and gives every vector a unique coordinate representation. If you switch from one basis to another, you need a rule that turns old coordinates into new coordinates. That rule is the change of basis matrix.
To build it, write each vector in the new basis as a linear combination of the old basis vectors. The coefficients become the columns of the matrix. So if you want to convert coordinates from basis A to basis B, you construct a matrix that tells you how the B-basis vectors look in A-coordinates. Multiplying that matrix by a coordinate vector in basis A gives the vector’s coordinates in basis B.
This is where a lot of confusion happens: the matrix does not move the actual geometric vector around in space. It only changes the description. The same arrow can have different coordinate vectors depending on whether you measure it using the standard Cartesian coordinates or some other basis adapted to the problem.
Because both bases describe the same vector space, the change of basis matrix is invertible. Its inverse reverses the process and sends B-coordinates back to A-coordinates. That matches the idea that you can always go back and forth as long as both sets are actually bases.
In this course, change of basis shows up anytime you want a cleaner coordinate system for a problem. For example, a matrix or linear transformation may look simpler in a basis made from eigenvectors, even though the underlying transformation has not changed. The basis changes the lens, not the math object itself.
Change of basis matrices connect the abstract side of linear algebra to the coordinate calculations you actually do. They show why a vector space is not tied to one fixed set of axes, and they make it possible to move between different descriptions of the same data without losing information.
That matters most when a problem is easier in one basis than another. In a standard basis, a transformation might look messy, but in a basis chosen to match the structure of the problem, the matrix can become simpler to work with. This is the same idea behind using eigenvectors in diagonalization, where a smart basis can turn repeated matrix multiplication into a much easier process.
Change of basis also sharpens your understanding of dimension and coordinate systems. If you know a space has dimension 3, you know any basis for that space has 3 vectors, but the actual coordinate triples can look completely different from one basis to another. The matrix records that translation.
In differential equations, this mindset shows up when systems are rewritten in forms that are easier to solve or interpret. Even when you are not explicitly building a change of basis matrix every time, the same logic is at work whenever you choose coordinates that simplify the problem.
Keep studying Linear Algebra and Differential Equations Unit 3
Visual cheatsheet
view galleryBasis
A basis is the starting point for any change of basis matrix. You cannot switch coordinates unless you know the vectors that define each coordinate system. The matrix is built from one basis written in terms of another, so understanding what makes a set of vectors a basis is the first step.
Coordinate Vector
A coordinate vector is what the change of basis matrix acts on. It is not the geometric vector itself, just the list of numbers that describes that vector in a chosen basis. When you switch bases, you are changing that list, not the underlying vector.
Linear Transformation
A change of basis matrix is a specific linear transformation between coordinate representations. It preserves linear structure while converting one description into another. That makes it a good example of how matrices can represent rules for moving between different viewpoints in a vector space.
Isomorphism
Different bases give different coordinate systems, but they describe the same vector space. That relationship is an isomorphism, meaning the spaces are structurally the same even if their coordinates look different. Change of basis is one of the clearest examples of that idea in action.
A problem set or quiz question will usually ask you to find a change of basis matrix, use it to convert coordinates, or identify the inverse matrix that switches back. The move is straightforward but easy to mix up: first decide which basis you are converting from and which basis you are converting to, then build the matrix with the correct columns in the correct order. If the question gives you a vector in one basis, you multiply by the change of basis matrix to rewrite it in the new basis. If it gives you the matrix in the opposite direction, use the inverse instead. A common mistake is treating the matrix like it changes the vector itself. On the page, the vector stays the same and only the coordinate description changes. In a longer assignment, you may also be asked to explain why a transformation looks simpler in a different basis, especially when the basis is chosen to match eigenvectors or another special structure.
A linear transformation is a rule that sends vectors to vectors. A change of basis matrix does not create a new geometric action on the vector space, it rewrites coordinates for the same vector in a different basis. They both use matrices, but they are doing different jobs.
A change of basis matrix converts coordinate vectors from one basis to another, but it does not change the vector itself.
You build the matrix by expressing the new basis vectors in terms of the old basis vectors, then using those coefficients as columns.
The matrix is invertible because you can always convert coordinates back if both sets of vectors are bases for the same space.
This idea is really about changing the lens on a vector space, not changing the space itself.
In this course, change of basis becomes especially useful when a different basis makes a matrix or system easier to work with.
They are matrices that convert a vector’s coordinates from one basis to another. The vector stays the same, but its coordinate representation changes because the basis vectors you are using to measure it have changed. This is a coordinate conversion tool, not a new transformation of the vector itself.
Write each vector in the target basis as a linear combination of the vectors in the starting basis. Put those coefficient lists into the columns of a matrix. If you need to go the other direction, take the inverse of that matrix.
Not exactly. A linear transformation changes vectors according to a rule, while a change of basis matrix rewrites the coordinates of the same vector in a different basis. Both use matrices, which is why they can look similar at first.
It lets you pick the coordinate system that makes the work cleaner. That shows up when a matrix becomes simpler in another basis, or when a differential equation system is easier to analyze after rewriting it in a better coordinate frame.