Transition state search

[isayev]

Kudos to the author, finally a descent open source python implementation! Quick question, can you do TS (saddle points) optimization out of the box?

[jhrmnn]

No, it can't. Most of the ingredients are here though. One would need to (1) remove the line interpolation and (2) modify the quadratic step such that (ii) the second instead of the first eigenvector of the RFO matrix is taken and (ii) the root of the shifted RFO eq is searched for between the first and second eigenvalue instead of below the first.

Also, while I'm sure that the algorithm used here is pretty close to the state of the art for minimization, I have no idea whether that would be also true for the TS search.

I'll leave this issue open as a feature request.

[espottesmith]

Hello.

Is this feature under active development, or are you still waiting for assistance? A native Python transition-state optimization algorithm would be greatly helpful for my research, and I would be interested in working to develop such an algorithm on top of the existing pyberny code.

[jhrmnn]

It is not under active development, and the status hasn't changed since my comment above.

I would most welcome this kind of effort and would be very happy to help if you encounter any issues or ambiguities with the existing code.

Also, I'm not aware that anyone else would be working towards this feature in Pyberny.

[espottesmith]

Excellent! I'll begin work and let you know what problems, if any, I encounter.