Cannot locate DDI on Linux

[ruslankotl]

Use of Gamess wrapper impossible on Linux machines due to chemtools providing an invalid command to the shell resulting in a DDI error and failed computation. Could you rework get_command module to provide a proper command as in pygamess:
exec_string = "{0} {1} {2} -ddi {3} {3} {4} -scr {5} > {6}".format(ddikick, gamess, self.jobname, self.num_cores, hostname, self.tempdir, self.gamout)
where ddikick is a path to DDI, gamess is a path to executable, self.jobname is an input file, num_cores is a number of cores, self.tempdir is a scratch folder, and self.gamout is an output file.

[lmmentel]

Hi @ruslankotl Thanks for taking the time to report this. Could you paste the actual error you're getting and a minimal example I could reproduce?

[ruslankotl]

Thank you for getting back @lmmentel

So I was following your [https://chemtools.readthedocs.io/en/latest/notebooks/GamessUS_tutorial.html](tutorial for Gamess(US)) using a new version Jupyter Notebook with Ubuntu 20.04 LTS installed on WSL.

Setting up the calculator was reasonably easy:

gamess = GamessUS(executable="/home/ruslan/gamess/gamess.01.x", version="01", runopts=["1"], scratch="/home/ruslan/scr")

rungms and ddikick methods have pointed to the correct locations, however your original method with specifying exevar instead of execultable did not work (should I have set up .bashrc?)

I then followed the instructions further, however instead of the result I have got the following message:

DDI Process 0: error code 911
DDI: Invalid command-line arguments!
Are you using the ddikick program??

Running the input file through the terminal directly worked normally.

[lmmentel]

I tried the tutorial you mention with the latest version of gamess and all works fine. It seems to me that you might have some configuration issues related to your gamess installation. To verify that you gamess installation runs OK try executing some examples directly from the terminal using rungms comman, if that works chemtools should work as well.

[ruslankotl]

I tried the tutorial you mention with the latest version of gamess and all works fine. It seems to me that you might have some configuration issues related to your gamess installation. To verify that you gamess installation runs OK try executing some examples directly from the terminal using rungms comman, if that works chemtools should work as well.

That's the thing: executing the example with chemtools yields a DDI error, but executing the example with ! ~/gamess/rungms h2_test1.inp 01 4 >& h2_test1.log within the same Jupyter Notebook proceeds normally.

[lmmentel]

Could you try using the path to your rungms as executable? I guess it should look something like

gamess = GamessUS(executable="/home/ruslan/gamess/rungms", version="01", runopts=["1"], scratch="/home/ruslan/scr")

and try again?

[ruslankotl]

Seemed to work, thank you very much!