Addressing Formulas for Elements of 3D Sparse Matrices in 1D Arrays

3D sparse matrices play significant role in many modern information retrieval methods. A huge number of computations with such matrices are performed by these methods, like clustering, latent semantic indexing. Since many of the elements in such matrices contain zeroes, thus a need to save space arises, either in memory or disk. Thus their storage in memory or disk should be very carefully designed. To save the space in memory or disk, efficient storage of elements and their retrieval is required. This paper explains two types of 3D sparse matrices known as lower triangular and upper triangular. The basic concept behind this paper is to calculate the formulas for accessing the address of elements of 3D sparse matrices in 1D array using row major and column major.