What it means
The vectors v₁, v₂, …, v_k are linearly independent if
- c₁v₁ + c₂v₂ + … + c_kv_k = 0 only when c₁ = c₂ = … = c_k = 0
If some combination with a coefficient other than 0 gives the zero vector, the vectors are linearly dependent.
That is the same as one of them being a combination of the others: from 2v₁ − v₂ − v₃ = 0 follows
v₃ = 2v₁ − v₂. Two vectors are dependent exactly when they are parallel, and three vectors in space are
dependent when they lie in one plane through the origin.
The determinant test
For n vectors in ℝⁿ, put them as the columns of an n×n matrix. The vectors are linearly independent exactly
when the determinant of that matrix is not 0. The card above is the determinant of three vectors u, v and w
in ℝ³, the determinant of a 3×3 matrix,
and it opens on the usual question: for which value of a coordinate are the vectors dependent? Give the
eight other coordinates and set the determinant to 0.
Holds when the number of vectors equals the dimension, so that the matrix is square. For another number of
vectors, see the last section.
Worked example
For which t are u = (1, 2, 3), v = (4, 5, 6) and w = (7, 8, t) dependent? With the vectors as columns,
- det = 1·5·t + 4·8·3 + 7·2·6 − 7·5·3 − 4·2·t − 1·8·6 = 5t + 96 + 84 − 105 − 8t − 48 = 27 − 3t
This is 0 for t = 9. Check: w = (7, 8, 9) is 2v − u = (8 − 1, 10 − 2, 12 − 3). For any other t, such as
t = 10 where the determinant is −3, the three vectors are independent.
The same matrix in the matrix calculator gives rank 2:
only two of the three columns are independent.
A second example
u = (1, 0, 2), v = (0, 1, 1) and w = (2, 3, 7). The determinant is 1·1·7 + 0 + 0 − 2·1·2 − 0 − 1·3·1 = 7 − 4 − 3 = 0.
The vectors are dependent, and the combination is w = 2u + 3v = (2, 0, 4) + (0, 3, 3).
Two vectors alone need no determinant. (1, 2, 3) and (2, 4, 6) are dependent since the second is twice the
first, and (1, 2, 3) and (2, 4, 7) are independent since neither is a multiple of the other.
Other numbers of vectors
| Vectors |
What decides |
| n vectors in ℝⁿ |
the determinant: independent exactly when it is not 0 |
| fewer than n vectors in ℝⁿ |
the rank of the matrix with the vectors as columns: independent exactly when it equals the number of vectors |
| more than n vectors in ℝⁿ |
always dependent |
| any set containing the zero vector |
always dependent |
The rank is found by Gaussian elimination or in
the matrix calculator. A basis of a space is a set of independent vectors that spans
it; see basis and dimension.
Common mistakes
- Testing the vectors in pairs. (1, 0, 0), (0, 1, 0) and (1, 1, 0) are pairwise non-parallel and still
dependent, since the third is the sum of the first two. Independence is about all combinations at once.
- Using the determinant when the matrix is not square. Three vectors in ℝ⁴ have no determinant. Use the
rank.
- Believing rows and columns give different verdicts. The determinant and the rank are the same for a
matrix and its transpose, so the vectors can stand as rows or as columns. The columns are needed only when
you want to read the combination itself off the null space.
Frequently asked questions
How many linearly independent vectors can there be in ℝ³?
At most three. Any four vectors in ℝ³ are dependent, and three independent vectors form a basis of ℝ³.
What does linearly dependent mean geometrically?
Two dependent vectors lie on one line through the origin. Three dependent vectors in space lie in one plane
through the origin, or on a line.
Is a single vector independent?
Yes, if it is not the zero vector. c·v = 0 with v ≠ 0 forces c = 0.