search

usnistgov / COSMOSAC

A Benchmark Implementation of COSMO-SAC
MIT License
51 stars 29 forks source link

About tutorial #26

Open Fxzhou0919 opened 1 year ago

Fxzhou0919 commented 1 year ago

Dear Prof. Bell I want to ask if there is a tutorial for reference?

ianhbell commented 1 year ago

Please check through the python and jupyter notebooks. There are quite a few examples.

Fxzhou0919 commented 1 year ago

Thank you! I also want to ask one more question. I used the Gaussian09 calculation to get my .cosmo file, and then call the to_sigma.py file. Unfortunately, i got the following error: 1665923007213 I also tested the .cosmo file in the /COSMOSAC/profiles/GAUSSIAN09_TEST directory and it works perfectly. I want to ask what's wrong with my cosmo file?

ianhbell commented 1 year ago

Is your input path correct? Pointing to the desired file? I don't see any error message.

rkrusushabh224 commented 1 year ago

I have the same issue. As I was trying to create a '.sigma' file for malonic acid, I got this message. Even though it is not an error, no '.sigma' file has been created. Screenshot 2023-06-13 182349 I have tried with both the averaging methods, I got the same output message. Refer to the image attached above.

ianhbell commented 1 year ago

Are you permitted to attach the file you tried to convert? Determining what input format you have has always been a bit fragile.

mim134 commented 4 months ago

Hey, I've been having the same issue from Gaussian files. It seems to be caused by line 555 in to_sigma.py assert(abs(sum(psigmaA)-meta['area [A^2]']) < 0.001)

The difference I'm getting for the molecule I'm testing is 0.0013. How important is this area check? I can generate a profile by removing this line but I'm concerned about the accuracy of the output. Thank you

ianhbell commented 4 months ago

I'm guessing you are just running into rounding errors. Are you able to post the output file?

mim134 commented 4 months ago

Here is the output file. Thanks! PFBA.txt

ianhbell commented 4 months ago

The area check is just a sanity check that things are more or less ok. Can you please email me your Gaussian output to ian.bell@nist.gov and I will take a quick glance?

mim134 commented 4 months ago

Sent! Thank you!