Dalton forum

Many thanks for pointing this out Alexander!

We will make a fix for the upcoming release early spring.

There are, and has been for quite some time, a plan to implement ecp's for LSDalton. This is in the making, but relies on developments in a sister library. There are no plans at current for X2C or ZORA, unfortunately.

You can still run all-electron calcluations of complexes with Pt and Au with LSDalton, but this is probably not a very good idea, as you will not account for any relatvisitic effects.

Dear Terri, The two-photon cross-sections for linear and polarized light are printed in the output under the line: ``` State No. Exc. Energy Linear pol. Circ. pol.``` Only the transition dipolmoments and Oscillator strengths are printed for the excitations. Permanent dipole moments for excited state...

PS:

1. you might want to add OpenMP (`--omp`) for better performance

2. we are now on the verge of making a new release (scheduled in November, ongoing work on branch `release/2018`)

3. please also note that it has been some time since we have tested the code using cray

Simen

Dear Alvaro,

Thanks for your feedback.

This problem should now have been fixed with the latest merge. The following should solve it:

```
git pull
git submodule update --init --recursive
cd build
make
```

Great if you can let us know if this solves it.

Best regards,
Simen

Dear Elias, Great to hear, and yes, ScaLapack should be used together with threaded BLAS. You can in fact also use the default ATOMS startguess together with the NOGCBASIS. I converges nicely with VanLenthe DIIS in 8 iteration for your 96-atom water cluster. So either replace H1DIAG with ATOMS or si...

Dear Elias, Clearly this is not the typical behaviour, and looking at your output it is clear that the PBLAS/ScaLapack interface is not working as it should for your run; since the CPU and wall timings of the matrix operations (in G_GRAD, AVERAG and G_DENS) are identical. So the first step here woul...

A new LSDalton version has been added to the download folder:

LSDalton1.3.tar.gz

This version includes:

1. bugfixes and improvements for the handling of non-integer point and molecular charges (as needed by Chemshell interface)
2. update of cmake structure and test handling

Ifort version 16 was released in 2016 and is not supported for dalton 2015. I can no longer remeber which versions we did support but I believe all 12, 13 and 14 versions. There is a short introduction to running lsdalton on the forum: http://daltonprogram.org/forum/viewtopic.php?f=15&t=61 The synta...

Dear Qinghua, There seems to be four different causes for errors here: 1. the dalton test case that fail is energy_lucita_gasci (the others are all lsdalton test cases). I don't know what causes this, but there is at least some indication in the log file 2. the dectests fails because they are only i...

Great, glad to hear.

Good luck!

He writes intel 2015. Intel 2015 is not supported for this release, and we recommend to use 2014 or earlier versions instead: http://dalton-installation.readthedocs.org/en/latest/doc/supported.html Technical: The problem you report is typically related to problems with inter-procedural optimisation,...

You can instead do a: export MATH_ROOT=/home/lz/TEST_BUILDS/atlas_gfortran and then run the setup: ./setup --fc=gfortran --cc=gcc --cxx=g++ --cmake=/home/lz/TEST_BUILDS/cmake-3.2.0-rc2-Linux-x86_64/bin/cmake This should automatically detect and add the needed libraries for you. With the explicit lib...

Good point! In the run directory each geometry is stored as MOLECULE.XXX, where XXX is the current geometry step. On convergence the final geometry is stored in MOLECULE.OUT. These files are in a format that can be used as input for a subsequent calculation. However, at current, I see no good way to...

Dear Varuniltr,

Apologies for the long response time on this one. Somehow the geometry step goes wrong, but I cannot really say why. If you still would like help to resolve this, please provide you input/output so that we can look into this.

Kind regards,
Simen

Hi Jeff,

I was confusing your error with another error, and my suggestion will of course not work. I think Radovan is right that it is an incompatibility between gnu and intel going on here (as it tries to use the gnu version of c++locale.h).

Best regards,
Simen

Hi Jeff,

For some reason it seems -lstdc++ is missing. I am trying to figure out why. If you are impatient I think:

./setup --fc mpif90 --cc mpicc --cxx mpicxx --mkl=parallel --mpi --explicit-libs "-L$MKLROOT -mkl -llstdc++"

might do the trick.

Best regards,
Simen

We have no a priori estimate of the memory consumption in LSDALTON. However, we do have an a posteriori printing of the memory statistics for a successful run, for example: Max allocated memory, TOTAL 43.272 GB Max allocated memory, type(matrix) 9.730 GB Max allocated memory, real(realk) 43.270 GB ....

The purpose of this tutorial is to get you acquainted with running LSDALTON. Before starting this tutorial you must first: 1. Successfully install lsdalton (full installation or lsdalton only) 2. Successfully complete the test suite (full or linsca only) See the post 'Installing Dalton' for installa...