Data analysis and mining in scientific domains involve storage of large-scale multi-dimensional datasets for scientific, statistical & engineering applications in multidimensional online analytical processing (MOLAP) databases. Because of the size of the datasets is increasing and the degree of data-sparsity is being high, it is important to find the suitable and efficient compression scheme for storing data at a minimal scheme. This paper represents a comparative analysis of Traditional and Modern Data Compression Schemes for Multi-Dimensional data ranging from dimension 1 to 3. The main idea is to compare the space savings of four different & significant compressions schemes i.e. Bit Map, Header Compression, Compressed Row Storage (CRS) & Extendible Array Based Compression Scheme (EaCRS). The results from experiments show that EaCRS scheme is better than the other schemes in case of space complexity especially for higher data density.