very different results from the notebook

theislab / cpa

The Compositional Perturbation Autoencoder (CPA) is a deep generative framework to learn effects of perturbations at the single-cell level. CPA performs OOD predictions of unseen combinations of drugs, learns interpretable embeddings, estimates dose-response curves, and provides uncertainty estimates.

BSD 3-Clause "New" or "Revised" License

83 stars 17 forks source link

very different results from the notebook #3

Closed bitcometz closed 1 year ago

bitcometz commented 2 years ago

hello，

I am trying to redo the notebook：

https://cpa-tools.readthedocs.io/en/latest/tutorials/GSM.html

But I got very different results from the notebook using the same data as inputs and the same code, For example:

my results:

1655719646018_EAC31BF5-C6EA-43df-AF1D-3E3AF823AFC2

the notebook:

my results: 1655719653499_7A78F9F2-FB74-4ef4-8584-2727A08031CC

the notebook:

Could you help with this problem ? Thanks !!!

Naghipourfar commented 2 years ago

Hi @bitcometz,

Thanks for letting us know. I'm very sorry for the late response. I will publish the notebook with the new hyper-parameter setting for our experiments ASAP.

bitcometz commented 2 years ago

Thanks !!!

Naghipourfar commented 2 years ago

Hi @bitcometz,

Just released a new version for the package with slight changes in input arguments and API in general. Also the tutorial notebooks are updated with latest API and hyper-parameter configuration. So I believe you can reproduce the results just by running the notebooks. Thanks very much for your patience. Please let me know if the issue still exists.

Best, Mohsen