TS search

3 posts / 0 new
Last post
TS search

Hello everyone,

My TS search log file says :

LBFGS converged , starting with PRFO.
and then lists down convergence values.
and then later says TS mode is number 2 with eigenvalue ...

So what does TS mode is number 2 means ? I can't consider that as probable TS?
Also in my input file I had specified .00135 as tolerance value but I can no where see that mentioned in any of the Target values.

You can reply me whenever you have time .
Waiting for your reply,
Thank you


Snippet of log file

HDLC step and gradient:
rfo MaxStep: 0.28657 Target: 0.54000E-02 converged? no Component: 1
rfo RMSStep: 0.62341E-02 Target: 0.36000E-02 converged? no
rfo MaxGrad: 0.43512E-02 Target: 0.13500E-02 converged? no Component: 50
rfo RMSGrad: 0.94967E-04 Target: 0.90000E-03 converged? yes

Cartesian gradient:
rfo MaxGrad: 0.44469E-02 Target: 0.13500E-02 converged? n/a Component: 42
rfo RMSGrad: 0.19043E-03 Target: 0.90000E-03 converged? n/a
Hessian is being updated using the Powell update

Hessian has 1 negative eigenvalue(s)
Eigenvalues of the Hessian:
-0.15681E-02 0.35409E-03 0.12549E-02 0.22399E-02 0.24256E-02 0.38335E-02
0.45482E-02 0.50570E-02 0.54701E-02 0.69956E-02 0.72968E-02 0.77646E-02
0.88569E-02 0.90292E-02 0.99785E-02 0.10650E-01 0.11886E-01 0.14014E-01
0.14512E-01 0.16073E-01 0.18402E-01 0.19443E-01 0.21060E-01 0.21540E-01
0.22860E-01 0.24887E-01 0.26765E-01 0.28655E-01 0.32966E-01 0.36218E-01
0.36541E-01 0.39737E-01 0.42029E-01 0.42876E-01 0.44255E-01 0.48025E-01
0.50458E-01 0.51336E-01 0.53135E-01 0.56323E-01 0.58643E-01 0.60877E-01
0.62537E-01 0.69973E-01 0.71739E-01 0.73773E-01 0.75409E-01 0.81727E-01
0.83401E-01 0.87222E-01 0.92748E-01 0.93630E-01 0.10159 0.10588
0.11054 0.11599 0.12113 0.12615 0.12933 0.13784
0.15544 0.16187 0.17422 0.18452 0.19578 0.21563
0.24428 0.25545 0.29851 0.35730 0.36806 0.75475
0.85179 1.0026 1.0238 1.1246 1.1449 1.1878
Overlap of current mode 2 with previous mode is 0.963

TS mode is number 2 with eigenvalue 0.35409E-03
Lambda that maximises along TS modes = 0.00256
Lambda that minimises along all modes = -0.00391
P-RFO step has length 1.64536
Lambda that minimises along all modes = -0.00953

Why isn't the algoithm

Why isn't the algoithm considering TS mode 1 (eigenvalue is negative) whereas for TS mode 2 (eigenvalue is positive) ?

Log in or register to post comments