Input matrix must be double or logical. Third input must be double or logical. Subscripts must be the same data type (single, double, or integer). Use sparse(m,n) to preallocate an m-by-n sparse matrix, or sparse(i,j,v) to specify the (i,j) locations of nonzero entries v. Inputs must be 2-D. Matrix is too large to convert to linear index. Sparse matrix sizes must be nonnegative integer scalars less than MAXSIZE as defined by COMPUTER. Use HELP COMPUTER for more details.