Ziyuan Zhao

Duplicated forces with multiple forcefields?

4

This is probably a continuation of the issue closed [here](https://github.com/openmm/openmm/issues/2481). The current OpenMM manual [says that](http://docs.openmm.org/latest/userguide/application/05_creating_ffs.html#using-multiple-files) when using multiple XML forcefield files, > For other forces, the files are effectively...

Support for appending mode for hdf5 writer?

1

Hi, I'm working on a rather large dataset and would appreciate the ability to continuously append chunks of trajectory to a hdf5 file, much like what the HDF5 reporter does...

Support for time delay in simulation reporters?

3

We are currently studying the effect of some dynamic perturbation on our simulation system and would want to interleave several reporters to collect state and coordinates data from the simulation...

I just found this issue while using this util function in my project, as shown here: https://github.com/deepmind/jraph/blob/51f5990104f7374492f8f3ea1cbc47feb411c69c/jraph/_src/utils.py#L614-L619 Is there any particular reason for using numpy instead of jax.numpy? I will...

Molecule.from_polymer_pdb does not register double bond stereochemistry (E/Z)

3

**Describe the bug** If we apply `Molecule.from_polymer_pdb` on a peptide with amino acid which sidechain has a E/Z stereobond (curiously, almost all amino acids do not), the function will complain...

请问代码中在哪里调unlabelled data的比例？

8

您好，最近拜读了你们的文章，想在别的数据集上复现一下。然后我对inductive learning这部分比较感兴趣，看你们文章中意思是包含了所有train split里面没有按few shot learning标记上的数据。但是在看代码的时候，没有找到是在哪里切换supervised learning和inductive learning，所以想在这里问一下，谢谢了！

I've observed a bug in this function - if `n_windows = 1` due to a large window size, then np.squeeze will make `user_dataset_windowed[user_id][0]` a number with dimension 0 rather than...