20 March 2012 11:36:51 AM

SVD_TRUNCATED
  C++ version
  Demonstrate the use of the truncated or economy-size
  Singular Value Decomposition (SVD) for cases where
  the sizes of M and N are very different.

SVD_TRUNCATED_U_TEST
  M = 4
  N = 3

  A:

  Col:          0             1             2       
  Row

    0:     0.218418      0.415307      0.043829  
    1:     0.956318     0.0661187      0.633966  
    2:     0.829509      0.257578     0.0617272  
    3:     0.561695      0.109957      0.449539  

  UN:

  Col:          0             1             2       
  Row

    0:    -0.183175     -0.646219     -0.706139  
    1:    -0.712493      0.427285    -0.0298313  
    2:    -0.507765      -0.58287      0.611171  
    3:    -0.448303      0.245135     -0.356298  

  SN:

  Col:          0             1             2       
  Row

    0:      1.58618             0             0  
    1:            0      0.488342             0  
    2:            0             0      0.276424  

  V:

  Col:          0             1             2       
  Row

    0:    -0.879083       -0.1604      0.448871  
    1:    -0.191193     -0.743962     -0.640286  
    2:    -0.436645      0.648685     -0.623337  

  Maximum error |A - U*S*V'| = 3.33067e-16

  Recomputed A = U * S * V':

  Col:          0             1             2       
  Row

    0:     0.218418      0.415307      0.043829  
    1:     0.956318     0.0661187      0.633966  
    2:     0.829509      0.257578     0.0617272  
    3:     0.561695      0.109957      0.449539  

SVD_TRUNCATED_V_TEST
  M = 3
  N = 4

  A:

  Col:          0             1             2             3       
  Row

    0:     0.218418      0.561695      0.257578      0.633966  
    1:     0.956318      0.415307      0.109957     0.0617272  
    2:     0.829509     0.0661187      0.043829      0.449539  

  U:

  Col:          0             1             2       
  Row

    0:    -0.458728     -0.880639     -0.118507  
    1:    -0.658087      0.426317     -0.620626  
    2:    -0.597068       0.20671        0.7751  

  SM:

  Col:          0             1             2       
  Row

    0:       1.4941             0             0  
    1:            0      0.681157             0  
    2:            0             0      0.366801  

  VM:

  Col:          0             1             2       
  Row

    0:    -0.819763      0.567879     0.0642131  
    1:    -0.381802     -0.446198     -0.744454  
    2:    -0.145029     -0.250892     -0.176649  
    3:    -0.401476     -0.644573      0.640671  

  Maximum error |A - U*S*V'| = 1.11022e-15

  Recomputed A = U * S * V':

  Col:          0             1             2             3       
  Row

    0:     0.218418      0.561695      0.257578      0.633966  
    1:     0.956318      0.415307      0.109957     0.0617272  
    2:     0.829509     0.0661187      0.043829      0.449539  

SVD_TRUNCATED
  Normal end of execution.

20 March 2012 11:36:51 AM