Concept

Kahan's matrix — where it appears

An upper triangular matrix with sⁱ down the diagonal and −c·sⁱ above it, on which every trailing column norm is exactly equal at every step. It is built from one parameter c in (0, 1) with s the root of 1 − c², it gives a greedy pivot rule nothing to choose between, and its smallest singular value falls far below the last diagonal entry a factorisation reports.

Named by 2 essays across one field — each of them below, with the objects they name alongside it.

Named alongside it

The objects these essays reach for when they reach for this one.

Column pivotingCounterexampleNumerical rankRank-revealing QRSingular valuesCondition-estimationEckart–YoungHouseholder reflectionLow-rank approximationLower boundQR factorisationSpectral decay

All concepts