Sparse matrix
From Wikipedia, the free encyclopedia
In the mathematical subfield of numerical analysis a sparse matrix is a matrix populated primarily with zeros.
Sparsity is a concept, useful in combinatorics and application areas such as network theory, of a low density of significant data or connections. This concept is amenable to quantitative reasoning. It is also noticeable in everyday life.
Huge sparse matrices often appear in science or engineering when solving problems for linear models.
When storing and manipulating sparse matrices on a computer, it is beneficial and often necessary to used specialized algorithms and data structures that take advantage of the sparse structure of the matrix. Operations using standard matrix structures and algorithms are slow and consume large amounts of memory when applied to large sparse matrices. Sparse data is by nature easily compressed, and this compression almost always results in significantly less memory usage. Indeed, some very large sparse matrices are impossible to manipulate with the standard algorithms.
Contents |
[edit] Storing a sparse matrix
The naive data structure for a matrix is a two dimensional array. Each entry in the array represents an element ai,j of the matrix and can be accessed by the two indices i and j. For a n×m matrix we need at least (n*m) / 8 bytes to represent the matrix when assuming 1 bit for each entry.
A sparse matrix contains many (often mostly) zero entries. The basic idea when storing sparse matrices is to only store the non-zero entries as opposed to storing all entries. Depending on the number and distribution of the non-zero entries, different data structures can be used and yield huge savings in memory when compared to a naive approach.
One example of such a sparse matrix format is the (old) Yale Sparse Matrix Format [1]. It stores an initial sparse N×N matrix M in row form using three arrays, A, IA, JA. NZ denotes the number of nonzero entries in matrix M. The array A then is of length NZ and holds all nonzero entries of M. The array IA stores at IA(i) the position of the first element of row i in the sparse array A. The length of row i is determined by IA(i+1) - IA(i). Therefore IA needs to be of length N + 1. In array JA, the column index of the element A(j) is stored. JA is of length NZ.
[edit] Example
A bitmap image having only 2 colors, with one of them dominant (say a file that stores a handwritten signature) can be encoded as a sparse matrix that contains only row and column numbers for pixels with the non-dominant color.
[edit] Diagonal matrices
A very efficient structure for a diagonal matrix is to store just the entries in the main diagonal as a one dimensional array. For n×n matrix we need only n / 8 bytes when assuming 1 bit for each entry.
[edit] Reducing bandwidth
The Cuthill-McKee algorithm can be used to reduce the bandwidth of a sparse symmetric matrix. There are however matrices for which the Reverse Cuthill-McKee algorithm performs better.
The National Geodetic Survey (NGS) uses Dr. Richard Snay's "Banker's" algorithm because on realistic sparse matrices used in Geodesy work it has better performance.
There are many other methods in use.
[edit] Definition
Given a sparse N×M matrix A the row bandwidth for the n-th row is defined as
- <math>b_n(\mathbf{A}) := \mathrm{min}_{1 \le m \le M} \lbrace m \mid a_{n, m} \neq 0 \rbrace </math>
The bandwidth for the matrix is defined as
- <math>B(\mathbf{A}) := \mathrm{max}_{1 \le n \le N} b_n(\mathbf{A})</math>
[edit] Reducing fill-in
The fill-in of a matrix are those entries which change from an initial zero to a non-zero value during the execution of an algorithm. To reduce the memory requirements and the number of arithmetic operations used during an algorithm it is useful to minimize the fill-in by switching rows and columns in the matrix. The symbolic Cholesky decomposition can be used to calculate the worst possible fill-in before doing the actual Cholesky decomposition.
There are other methods than the Cholesky decomposition in use. Orthogonalization methods (such as QR factorization) are common, for example, when solving problems by least squares methods. While the theoretical fill-in is still the same, in practical terms the "false non-zeros" can be different for different methods. And symbolic versions of those algorithms can be used in the same manner as the symbolic Cholesky to compute worst case fill-in.
[edit] Solving sparse matrix equations
Both iterative and direct methods exist for sparse matrix solving.
[edit] See also
[edit] References
- Sparse Matrix Multiplication Package, Randolph E. Bank, Craig C. Douglas [1]
- Pissanetzky, Sergio 1984, "Sparse Matrix Technology", Academic Press
- R. A. Snay. Reducing the profile of sparse symmetric matrices. Bulletin Géodésique, 50:341–352, 1976. Also NOAA Technical Memorandum NOS NGS-4, National Geodetic Survey, Rockville, MD.
[edit] Further reading
- Norman E. Gibbs, William G. Poole, Jr. and Paul K. Stockmeyer (1976). "A comparison of several bandwidth and profile reduction algorithms". ACM Transactions on Mathematical Software 2 (4): 322–330.
- John R. Gilbert, Cleve Moler and Robert Schreiber (1992). "Sparse matrices in MATLAB: Design and Implementation". SIAM Journal on Matrix Analysis and Applications 13 (1): 333–356.
- Sparse Matrix Algorithms Research at the University of Florida, containing the UF sparse matrix collection.cs:Řídká matice
