Polar Decomposition
M = UP: U unitary (rotation/reflection), P positive semidefinite (scaling). Analog of z = re^(iθ) for complex numbers. P = √(M*M), U = MP⁻¹.
Did our AI summary help? Let us know.
P = √(MᵀM) unique positive semidefinite. U orthogonal (real) or unitary (complex). Like polar form z = r·e^(iθ).
Ready to run the numbers?
Why: Polar decomposition separates rotation from scaling. Used in continuum mechanics, computer graphics, and matrix analysis.
How: P = √(MᵀM) (positive semidefinite square root). U = MP⁻¹. When M invertible, U unique. P = (MᵀM)^(1/2).
Run the calculator when you are ready.
Matrix Size
Example Matrices
Identity
For the identity matrix, U = I and P = I
Rotation
For rotation matrices, U = the rotation matrix and P = I
Symmetric
For symmetric matrices, U = I and P = the symmetric matrix
Matrix Input
What is Polar Decomposition?
Polar decomposition is a matrix factorization technique that decomposes a square matrix M into the product of a unitary matrix U and a positive semidefinite Hermitian matrix P. This factorization is analogous to the polar form of a complex number (z = reiθ), where a complex number is represented as the product of its magnitude and a unit complex number.
Key Properties:
- Every square matrix has a polar decomposition M = UP
- U is a unitary matrix (U*U = UU* = I)
- P is a positive semidefinite Hermitian matrix
- The decomposition is unique when M is non-singular
Applications
Computer Graphics
In computer graphics, polar decomposition separates rotation from scaling/shearing, enabling smooth interpolation between transformations.
Mechanical Engineering
Used in continuum mechanics to separate rigid body rotation from pure deformation in the analysis of material deformations.
Quantum Physics
Helps analyze density matrices and quantum transformations, providing insights into quantum state evolution.
Signal Processing
Used in phase retrieval problems and filter design, separating magnitude and phase information.
Calculation Methods
Several methods exist for computing the polar decomposition:
SVD Method
Using Singular Value Decomposition (SVD): If M = WΣV*, then U = WV* and P = VΣV*. This is the most stable method, especially for ill-conditioned matrices.
Newton's Method
An iterative approach where Uk+1 = 0.5(Uk + (Uk-1)*), which converges quadratically for well-conditioned matrices.
Frequently Asked Questions
What's the difference between polar decomposition and eigendecomposition?
Eigendecomposition expresses a matrix using its eigenvalues and eigenvectors (M = PDP-1), revealing its intrinsic dynamics. Polar decomposition (M = UP) separates rotation from scaling, revealing its geometric action. Eigendecomposition only exists for diagonalizable matrices, while polar decomposition exists for all square matrices.
Can polar decomposition be computed for singular matrices?
Yes, polar decomposition exists for all square matrices, including singular ones. However, for singular matrices, the unitary part U is not unique in the null space of the matrix. The SVD-based method provides a standard choice for the unitary part.
Related Calculators
For educational and informational purposes only. Verify with a qualified professional.
🧮 Fascinating Math Facts
U preserves lengths
P has nonnegative eigenvalues
Related Calculators
L U Decomposition Calculator
L U Decomposition Calculator - Calculate and learn about linear-algebra concepts
MathematicsQ R Decomposition Calculator
Q R Decomposition Calculator - Calculate and learn about linear-algebra concepts
MathematicsCholesky Decomposition Calculator
Cholesky Decomposition Calculator - Calculate and learn about linear-algebra concepts
MathematicsSVD Calculator
S V D Calculator - Calculate and learn about linear-algebra concepts
MathematicsPseudoinverse Calculator
Pseudoinverse Calculator - Calculate and learn about linear-algebra concepts
MathematicsAdjoint Matrix Calculator
Adjoint Matrix Calculator - Calculate and learn about linear-algebra concepts
Mathematics