Running MRCC optimizations in a stepwise fashion

And sorry to add another step to my quandary:
If the lambda equations do not converge, is there a way to restart the lambda equations? It looks like my new job has reset to the initial guess for the lambda equations.

Dear Nathan,

The program always calculates the density matrix. If you don't like it, you can turn it off. Edit file mrcc.f, search for "Convergence not achieved", and insert "stop" after that line.

The success message for the density matrix calculation is "Total energy from RDM ...". Thereafer the program writes the density to an interface file for Cfour, which takes for a while.

Concening your last message: didn't you forget to set rest=1 in the input file? If the lambda equations do not converge and rest=1, the initial guess will not be reset but the program will restart from the old amplitudes.

Mihaly,
I will modify the code so that it does not calculate the density matrix if the lambda equations do not converge. But it still looks like calculation of the density matrix is stalling somewhere if the lambda equations are not converged.

I am using rest=1
The CC equations successfully restart, but the lambda equations restart from the initial guess. A bug perhaps?

Also, are you sure I should put "stop" after the "Convergence not achieved" text? Why not add the safer:
call mrccend(1)

Dear Nathan,
I cannot reproduce this. Could you prepare a sample input file for a short job?

Both of them are OK.

rayultine wrote: Also, are you sure I should put "stop" after the "Convergence not achieved" text? Why not add the safer:
call mrccend(1)