[BUG] Failure in Gaussian's parser

[Zhang-pchao]

After Gaussian has calculated all the cluster candidates, it terminates in step 8.
run.log reports an error ValueError: could not convert string to float: '219.870882556-1030.493229556'.

If other documents are needed, I can provide them in time.
run.log

[AnguseZhang]

@njzjz Would you please see this issue ?

[njzjz]

@njzjz Would you please see this issue ?

I am studying what ASE does.

[njzjz]

@Zhang-pchao could you post the Gaussian log files that can be a test?

[Zhang-pchao]

@Zhang-pchao could you post the Gaussian log files that can be a test?

output_pm6d3.log

[Zhang-pchao]

@Zhang-pchao could you post the Gaussian log files that can be a test?

output_pm6d3.log

I found that if I change "fp_task_max" from 1000 to 600, it can run. Could this bug be due to too many cluster candidates selected?

[njzjz]

I don't find 219.870882556-1030.493229556 in output_pm6d3.log.

[Zhang-pchao]

I don't find 219.870882556-1030.493229556 in output_pm6d3.log.
The reason for the bug: the overlapping of atoms in the clusters caused the inaccurate force of Gaussian calculations.
output.log

[njzjz]

This would be another issue.

…

On May 10, 2021, at 00:57, Zhang-pchao ***@***.***> wrote:  I don't find 219.870882556-1030.493229556 in output_pm6d3.log. The reason for the bug: the overlapping of atoms in the clusters caused the inaccurate force of Gaussian calculations. output.log — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or unsubscribe.

[njzjz]

@njzjz Would you please see this issue ?

I am studying what ASE does.

Oh, I found ASE also has this bug... Let me consider how to resolve it.