Incremental calculation of inverse of a matrix
Does there exist a fast way to calculate the inverse of an $N \times N$ matrix, if we know the inverse of the $(N-1) \times (N-1)$ sub-matrix?
For example, if $A$ is a $1000 \times 1000$ invertible matrix for which the inverse is known, and $B$ is a $1001 \times 1001$ matrix obtained by adding a new row and column to $A$, what is the best approach for calculating inverse of $B$?
Blockwise inverse