Solving for extra Excited States

Clicked A Few Times

8:01:33 AM PDT - Tue, Sep 25th 2012
Hello, I am wondering if there is a way to solve more roots in a TCE calculation without needing to resolve for the ones already computed. So for example if a set NROOTS 5 , and after the calculation is finished i was not able to locate my excited state of interest, and need to run another calculation and solve for 5 additional roots. Can we read the first 5 solved roots from the previous calculation and just solve for the additional 5, or all the roots will have to be resolved. This is particularly valuable for coupled clusters calculations. Thanks in advance.

Clicked A Few Times

11:57:03 AM PDT - Mon, Oct 1st 2012
Unfortunately, we do not have this functionality implemented. So, it means that if you want the root #7 you need to ask for 7 roots. The best solution to this problem is to track the root of interest only. However, this requires a good starting vectors (usually of better quality than the CIS vectors used in the current implementation). I will be trying to utilize our active-space EOMCCSd approach for this purpose. I should have a working version of new solver available next year. Best, Karol