Given a rectangular non-negative matrix X#95;0 and two positive marginals a and b, the algorithm generates a sequence of matrices (X#95;n) starting at X#95;0, supposed to converge to a biproportional fitting, that is, to a matrix Y whose marginals are a and b and of the form Y#61;D#95;1X#95;0D#95;2, for some diagonal matrices D#95;1 and D#95;2 with positive diagonal entries. When a biproportional fitting does exist, it is unique and the sequence (X#95;n) converges to it at an at least geometric rate.
Source: wiktionary