学习QR分解最好的一句话是这样说的:“把质的困难转化为量的复杂,转化前是一个相对复杂的矩阵,转化后矩阵简单但是数量多于一个。”真的是精髓之所在。
Decomposition | Method called in Eigen | Condition satisfied by matrix A | Computation Speed | Computational Accuracy |
---|---|---|---|---|
PartialPivLU | partialPivLu() | Invertible | ++ | + |
FullPivLU | fullPivLu() | None | - | +++ |
HouseholderQR | householderQr() | None | ++ | + |
ColPivHouseholderQR | colPivHouseholderQr() | None | + | ++ |
FullPivHouseholderQR | fullPivHouseholderQr() | None | - | +++ |
LLT | llt() | Positive definite | +++ | + |
LDLT | ldlt() | Positive or negative semidefinite | +++ | ++ |
表格来源:https://my.oschina.net/VenusV/blog/3030201
QR相比LU针对的不再是方阵,求逆算法效率更高,具有数值稳定的优点,这种直接方法的数值稳定性比LU分解要好。Cholesky是LU的特殊情况。