emdgroup/baybe

Batch_size error

CatSci opened this issue ยท 10 comments

Hi,

I am getting this error "TypeError: Campaign.recommend() got an unexpected keyword argument 'batch_size'". I am attaching snapshot for your reference. Could you please help me to sort this error out because till yesterday it was working fine.

error

Hello @CatSci, thanks for your question!

Could you maybe specify how you used and installed the package? That is, did you install via pip, did you install the latest version directly from the repository and which python version are you using?

Also, some other user experienced this as well, and it turned out that the reason was in fact a de-sync between the installed version of the package itself and the executed files. So you might also want to have a look here #121 and see if this happens to be related to your issue.

Thanks for the quick response. I am using this command "pip install 'baybe[chem,simulation]'" and python version is 3.10. And i will have a look.

You're welcome ๐Ÿ˜„
I just tested installing BayBE with python 3.10 and executed some code that contains the campaign.recommend() call using the batch_size argument, and it worked for me. Also, we test our examples regularly, so my guess is that there is indeed some de-sync in your project. Just keep me updated on your investigation and results, happy to assist ๐Ÿ‘

Thanks, I am using campaign.from_json() and then using campaign.recommend(). I will keep you updated.

image

image

Hi @CatSci the keyword batch_size of recommend was just recently introduced as renaming of the old keyword batch_quantity, please verify that you have a version > 0.7.2 installed since this deprecation happened there. Can you confirm which baybe version exactly is installed?

Also, since the old keyword was batch_quantity and our recommend function still has a batch_quantity keyword (for deprecation reasons) I would prob also try and rename the keyword batch_quantity to something else in your scripts to be absolutely 100% safe

Thanks, I will check the version and other recommendations from your side.

@CatSci did you manage to solve the error? If so, please feel free to close this issue :)

I am sorry I didn't close this earlier. Your suggestions helped to fix the issue. Thank you so much of helping me out

No problem and happy that we could help ๐Ÿ˜„