How To Find Basis For Eigenspace

Finding a basis for an eigenspace is a key skill in linear algebra because it shows you exactly which directions are stretched, compressed, or reversed by a linear transformation. Given a matrix (A) and one of its eigenvalues (\lambda), the eigenspace for (\lambda) is the set of all vectors (\mathbf{x}) that satisfy

[ A\mathbf{x}=\lambda \mathbf{x}, \quad \mathbf{x}\neq \mathbf{0}. ]

A basis for an eigenspace is a set of linearly independent eigenvectors that spans that eigenspace. In practice, you find it by solving a homogeneous system of linear equations.

Introduction to Eigenspaces

An eigenvector is a nonzero vector that does not change direction when multiplied by a matrix. Instead, it is only scaled by a number called an eigenvalue. If

[ A\mathbf{v}=\lambda \mathbf{v}, ]

then (\mathbf{v}) is an eigenvector of (A) corresponding to the eigenvalue (\lambda).

The eigenspace for (\lambda) includes all eigenvectors associated with (\lambda), along with the zero vector. Even though eigenvectors themselves must be nonzero, the eigenspace as a subspace includes (\mathbf{0}) Simple as that..

To find a basis for an eigenspace, you are really asking:

Which vectors satisfy (A\mathbf{x}=\lambda \mathbf{x})?

Rearranging the equation gives:

[ A\mathbf{x}-\lambda \mathbf{x}=\mathbf{0} ]

or

[ (A-\lambda I)\mathbf{x}=\mathbf{0}. ]

So, to find a basis for the eigenspace, you solve the null space of (A-\lambda I).

Step 1: Find the Eigenvalues

Before finding a basis for an eigenspace, you need to know which eigenvalue you are working with. If you already have an eigenvalue, you can skip this step. Otherwise, begin by solving the characteristic equation:

[ \det(A-\lambda I)=0. ]

Here:

• (A) is the matrix,
• (I) is the identity matrix of the same size,
• (\lambda) represents the eigenvalues.

The solutions to this equation are the eigenvalues of (A).

Here's one way to look at it: suppose

[ A= \begin{bmatrix} 4 & 1 \ 2 & 3 \end{bmatrix}. ]

Then

[ A-\lambda I= \begin{bmatrix} 4-\lambda & 1 \ 2 & 3-\lambda \end{bmatrix}. ]

The determinant is

[ \det(A-\lambda I)=(4-\lambda)(3-\lambda)-2. ]

Simplifying:

[ (4-\lambda)(3-\lambda)-2 =12-7\lambda+\lambda^2-2 =\lambda^2-7\lambda+10. ]

So:

[ \lambda^2-7\lambda+10=0. ]

Factoring:

[ (\lambda-5)(\lambda-2)=0. ]

That's why, the eigenvalues are:

[ \lambda=5 ]

and

[ \lambda=2. ]

Step 2: Choose an Eigenvalue

Once you have the eigenvalues, choose one eigenvalue at a time. To give you an idea, let us find a basis for the eigenspace corresponding to

[ \lambda=5. ]

The eigenspace is the set of all vectors (\mathbf{x}) such that:

[ (A-5I)\mathbf{x}=\mathbf{0}. ]

This means we subtract (5) from each diagonal entry of (A).

[ A-5I= \begin{bmatrix} 4-5 & 1 \ 2 & 3-5 \end{bmatrix}

\begin{bmatrix} -1 & 1 \ 2 & -2 \end{bmatrix}. ]

Step 3: Solve the Homogeneous System

Now solve:

[ \begin{bmatrix} -1 & 1 \ 2 & -2 \end{bmatrix} \begin{bmatrix} x_1 \ x_2 \end{bmatrix}

\begin{bmatrix} 0 \ 0 \end{bmatrix}. ]

This gives the equations:

[ -x_1+x_2=0 ]

and

[ 2x_1-2x_2=0. ]

Both equations simplify to:

[ x_1=x_2. ]

So every eigenvector for (\lambda=5) has the form:

[ \mathbf{x}= \begin{bmatrix} x_1 \ x_2 \end{bmatrix}

\begin{bmatrix} t \ t \end{bmatrix}

t \begin{bmatrix} 1 \ 1 \end{bmatrix}, ]

where (t\neq 0) for eigenvectors.

A basis for this eigenspace is therefore:

[ \left{ \begin{bmatrix} 1 \ 1 \end{bmatrix} \right}. ]

This means the eigenspace is the line through the origin in the direction of (\begin{bmatrix}1\1\end{bmatrix}) Not complicated — just consistent. Simple as that..

Step 4: Repeat for the Other Eigenvalue

Now find the eigenspace for

[ \lambda=2. ]

Compute:

[ A-2I= \

$ A-2I= \begin{bmatrix} 4-2 & 1 \ 2 & 3-2 \end{bmatrix}

\begin{bmatrix} 2 & 1 \ 2 & 1 \end{bmatrix}. $

Step 4a: Solve the Homogeneous System

Now solve:

$ \begin{bmatrix} 2 & 1 \ 2 & 1 \end{bmatrix} \begin{bmatrix} x_1 \ x_2 \end{bmatrix}

\begin{bmatrix} 0 \ 0 \end{bmatrix}. $

This gives the equations:

$ 2x_1+x_2=0 $

and

$ 2x_1+x_2=0. $

Both equations are identical, so they simplify to:

$ x_2=-2x_1. $

Every eigenvector for $\lambda=2$ has the form:

$ \mathbf{x}= \begin{bmatrix} x_1 \ x_2 \end{bmatrix}

\begin{bmatrix} x_1 \ -2x_1 \end{bmatrix}

x_1 \begin{bmatrix} 1 \ -2 \end{bmatrix}, $

where $x_1\neq 0$ for eigenvectors.

A basis for this eigenspace is therefore:

$ \left{ \begin{bmatrix} 1 \ -2 \end{bmatrix} \right}. $

This means the eigenspace is the line through the origin in the direction of $\begin{bmatrix}1\-2\end{bmatrix}$ Not complicated — just consistent..

Summary of Results

For the matrix

$ A= \begin{bmatrix} 4 & 1 \ 2 & 3 \end{bmatrix}, $

we found two eigenvalues and their corresponding eigenspaces:

Each eigenspace is one-dimensional, meaning each eigenvalue has exactly one linearly independent eigenvector (up to scalar multiples). The two basis vectors $\begin{bmatrix}1\1\end{bmatrix}$ and $\begin{bmatrix}1\-2\end{bmatrix}$ are linearly independent, which confirms that the matrix $A$ is diagonalizable.

Conclusion

Finding a basis for an eigenspace is a systematic process that relies on three key ideas. First, you determine the eigenvalues by solving the characteristic equation $\det(A-\lambda I)=0$. Even so, second, for each eigenvalue, you form the matrix $A-\lambda I$ and find its null space by solving the homogeneous system $(A-\lambda I)\mathbf{x}=\mathbf{0}$. Third, the non-trivial solutions to that system form the eigenvectors, and a maximal set of linearly independent eigenvectors constitutes a basis for the eigenspace Less friction, more output..

This procedure generalizes to matrices of any size. For larger matrices, the null space may have dimension greater than one, yielding multiple basis vectors per eigenvalue. So naturally, in such cases, the same row-reduction technique applies — you simply identify all free variables and express each in terms of the others to obtain a complete basis. Understanding eigenspaces is fundamental to many applications in linear algebra, including diagonalization, spectral decomposition, and the analysis of dynamical systems.

Most guides skip this. Don't.