r*******y
发帖数: 1081 | |
a*******h
发帖数: 123 | |
S*********g
发帖数: 5298 | 3 试试gsl
【在 r*******y 的大作中提到】
: 该用哪个库，谢谢
|
y****e
发帖数: 23939 | |
n**f
发帖数: 121 | 5 How do you build Lapack?
I know binaries exist for many operating systems. But I am still curious how
you build it from source, since it might be safer. Do you go through the
f2c -> gcc or other compiler path?
【在 a*******h 的大作中提到】
: 不能。用 lapack。
|
n**f
发帖数: 121 | 6 I thought BLAS only comes with basic operations...and not necessarily good
ones for both dense and sparse.
【在 y****e 的大作中提到】
: blas 和 lapak
|
n**f
发帖数: 121 | 7 Depending on how big/sparse the matrice are, and how frequently you have to
solve this, you may want to try LU or SVD using numerical recipe.
【在 r*******y 的大作中提到】
: 该用哪个库，谢谢
|
p***o
发帖数: 1252 | 8 For LU, SuperLU is a good library.
to
【在 n**f 的大作中提到】
: Depending on how big/sparse the matrice are, and how frequently you have to
: solve this, you may want to try LU or SVD using numerical recipe.
|
n**f
发帖数: 121 | 9 How about its capability to handle singularity and near-singularity?
【在 p***o 的大作中提到】
: For LU, SuperLU is a good library.
:
: to
|
p***o
发帖数: 1252 | 10 Better check their documentation and publication.
My systems are quite nice.
【在 n**f 的大作中提到】
: How about its capability to handle singularity and near-singularity?
|
|
|
n**f
发帖数: 121 | 11 how big are your matrices?
【在 p***o 的大作中提到】
: Better check their documentation and publication.
: My systems are quite nice.
|
p***o
发帖数: 1252 | 12 <10K to a few M.
【在 n**f 的大作中提到】
: how big are your matrices?
|
S*********g
发帖数: 5298 | 13 试试这个gsl？
http://www.gnu.org/software/gsl/
【在 p***o 的大作中提到】
: <10K to a few M.
|
p***o
发帖数: 1252 | 14 I am talking about large sparse matrix.
【在 S*********g 的大作中提到】
: 试试这个gsl？
: http://www.gnu.org/software/gsl/
|
t*****n
发帖数: 4908 | 15 Eigen c++的矩阵库，效率还不错。跟MKL有的一比。纯头文件，即编即用。 |
t*****n
发帖数: 4908 | 16 UMFPACK + GOTOBLAS2
解稀疏矩阵最好了。
【在 p***o 的大作中提到】
: I am talking about large sparse matrix.
|
c***r
发帖数: 1570 | |
c***r
发帖数: 1570 | 18 for the large sparse matrix, try petsc |
