CS5107

error name needs a square symmetric matrix in this version.

#What it means

In this version, matrix.eigenvalues and matrix.eigenvectors work on square symmetric matrices only: the value at row i, column j must equal the value at row j, column i.

#Typical cause

• A matrix that is not symmetric, such as one built from unrelated data.

#How to fix it

Make sure the matrix is symmetric, for example a covariance matrix, or build it as (m + m.transpose()) / 2 when that suits your purpose.

#Example

//@chartscript 1
indicator("Eigenvalues")
const m = matrix.new<float>(2, 2, 1.0)
m.set(0, 1, 2.0)
const ev = m.eigenvalues()
plot(ev.get(0))

Fixed

//@chartscript 1
indicator("Eigenvalues")
const m = matrix.new<float>(2, 2, 1.0)
m.set(0, 1, 2.0)
m.set(1, 0, 2.0)
const ev = m.eigenvalues()
plot(ev.get(0))

#Related

• CS5106 — The matrix sizes do not fit (detail).