Abstract: Techniques are described for compressing multidimensional datasets. For instance, a compression module obtains lists of indices of data along dimensions of the multidimensional dataset. The compression module calculates lists of offset values from the lists of coordinates. The compression module then performs a statistical redundancy data compression algorithm on the lists of offset values. Because the lists of offset values may contain more repetitious values than the initial lists of coordinates, a compression ratio on the multidimensional dataset may be higher than without first calculating the lists of offset values.