1
$\begingroup$

I am attempting to complete calculations for atoms using CFOUR.
I cloned a GitHub repository that included a correct GENBAS file.
I then created a folder outside of the directory of the git clone to complete the calculations.
In this folder, I added a correct ZMAT input, a correct script to run the CFOUR calculation, and, most importantly, a symbolic link to the GENBAS file in the git clone directory.

The calculation did not yield a working output.
I attempted to change the ZMAT input, as I wondered whether that was the source of the issue.
I also tested different scripts to run the calculations.
An output was generated in all these cases however it never included a calculation as it couldn’t detect a basis-set in a GENBAS file.

I only resolved the problem by copying the GENBAS file directly into the directory where I was running the calculations (and removing the symlink).

Are symlinks therefore not compatible with CFOUR calculations?

Many thanks for your help.

$\endgroup$
8
  • $\begingroup$ You need to show what you tried, and what the result was. $\endgroup$ Commented Mar 29 at 14:51
  • 1
    $\begingroup$ Please clarify your specific problem or provide additional details to highlight exactly what you need. As it's currently written, it's hard to tell exactly what you're asking. $\endgroup$
    – Community Bot
    Commented Mar 29 at 20:27
  • $\begingroup$ Symlinks look like normal files to computer programs. There is no issue to use a symlink to a central GENBAS file in CFOUR calculations. You probably had a bad symlink. $\endgroup$ Commented Apr 1 at 9:40
  • $\begingroup$ @SusiLehtola How can a symlink be “bad”? I’ll update my post with more details. $\endgroup$
    – TheorVHP
    Commented Apr 2 at 13:04
  • 1
    $\begingroup$ Perhaps you have to be even more precise about what you did. What was the exact ln -s command? Was the file on a different file system? Where is your CFOUR input file and output file? $\endgroup$ Commented Apr 2 at 23:49

1 Answer 1

0
$\begingroup$

I believe I deduced the source of the issue.

When I originally created the symlink, I did so while in the directory containing the GENBAS file.
I therefore could use the following command:

ln -s .\GENBAS <path of the link to be created>

As a result, when I was in the directory with the symlink, it read the symlink as

symlink to —> .\GENBAS

as I verified using ls -l.
The 0submit bash script running the CFOUR calculations could therefore not find the GENBAS file as there was no such file with path .\GENBAS in this calculation directory.

To resolve the issue, I instead created the symlink with the command

ln -s <full path>/GENBAS <path of the link to be created>

which therefore resulted in an ls -l output of

symlink to —> <full path>\GENBAS

which the bash script could find.

I was fortunate that I saved my original symlink to compare the methods.

I can confirm that the second method yielded correct results.

$\endgroup$
10
  • $\begingroup$ Hello! It is great that you have made progress in answering your question! As a rule of thumb, stackexhange tries to keep each question singular. So it would probably be better if you either update the original question with the new question you now face, or accept this answer here and start a new question. $\endgroup$ Commented Apr 4 at 14:16
  • $\begingroup$ @AndreyPoletayev thanks for that comment. I think it would even be inappropriate to add new questions to this question post: I recommend for the user to create new questions posts for each new question that arises. Also, TheorVHP: Do you now see why I wrote comments twice on your question post, saying that you need to tell us precisely what you did? The original question was ill-posed, because symlinks to a GENBAS file do work in CFOUR. We needed to know what you did (and hence what your question was). $\endgroup$ Commented Apr 4 at 17:24
  • $\begingroup$ @AndreyPoletayev, thank you for the insight into the etiquette on MMSE. I’m a new user so any information is helpful. $\endgroup$
    – TheorVHP
    Commented Apr 5 at 11:20
  • $\begingroup$ @Ihavenofreetimeanymore, thanks to you too. I’ll make another post in the relevant Stack Exchange about the question of copying symlinks. Yes, I now see why adding specific detail was required to solve the issue. $\endgroup$
    – TheorVHP
    Commented Apr 5 at 11:20
  • $\begingroup$ @TheorVHP it's okay to ask here. Your question about symlinks was answered here by Susi Lehtola quite quickly. $\endgroup$ Commented Apr 5 at 12:01

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .