One of the most useful features of PSA is its ability to extract Hausdorff pair (HP) and nearest neighbor (NN) pair information (frames and distances) between pairs of paths using hausdorff_neighbors(), allowing users to examine the atomistic determinants of path differences. However, the frame and distance information is not only complicated, but it is also a clumsy, time-consuming process to carry over this information into a separate visualization program like VMD where one must manually load the trajectories (and also possibly the extracted structure files corresponding to the HP or NN frames) in order to do exploratory HP/NN analysis.
The proposed solution is to provide an enhanced visualization backend based on nglview that permits a user to directly view HP/NN information in a Jupyter Notebook. A complete solution should use PSAIdentifier (see Issue #10) as a tool for extracting the correct trajectories from a similarity analysis performed by a user.
One of the most useful features of PSA is its ability to extract Hausdorff pair (HP) and nearest neighbor (NN) pair information (frames and distances) between pairs of paths using
hausdorff_neighbors(), allowing users to examine the atomistic determinants of path differences. However, the frame and distance information is not only complicated, but it is also a clumsy, time-consuming process to carry over this information into a separate visualization program like VMD where one must manually load the trajectories (and also possibly the extracted structure files corresponding to the HP or NN frames) in order to do exploratory HP/NN analysis.The proposed solution is to provide an enhanced visualization backend based on nglview that permits a user to directly view HP/NN information in a Jupyter Notebook. A complete solution should use
PSAIdentifier(see Issue #10) as a tool for extracting the correct trajectories from a similarity analysis performed by a user.