Feature request: analytic CCSD(T) gradients/hessians

I was wondering whether you were planning on ever adding the analytic gradients and/or hessians for CCSD(T) in MRCC natively? I am aware that MRCC can interface with CFOUR to use their gradients, but, based on my experience with CFOUR, MRCC is far quicker, parallelises more efficiently and is easier to obtain.

I know it's unlikely to happen but, based on the speed of your CCSDT calculations (which absolutely amazed me!) I thought I would ask.

Thank you for the appreciation and interest.

The development of an efficient CCSD(T) analytic gradient code is at an advanced stage, it will happen. Analytic Hessian is not in the short term plans, seminumerical should be fine for most purposes.

Best regards,
Peter

That's amazing, I assumed there were no plans to do it at all! Looking forward to it!!! 😃

The main reason for the analytic hessians would be for (G)VPT2 and other fun things, but I understand that can be very niche, considering it's generally so expensive. But, if your code is as fast as I think it is, CCSD(T) anharmonic frequencies could be far more practical than they are now... (CFOUR is far too slow for it to be worth the effort, usually.)