Added draft of vasp_util.py for VASP trajectories and OTF #82
Conversation
…ite and check unit tests
Codecov Report
@@ Coverage Diff @@
## master #82 +/- ##
==========================================
+ Coverage 50.97% 51.74% +0.76%
==========================================
Files 29 31 +2
Lines 4716 4791 +75
==========================================
+ Hits 2404 2479 +75
Misses 2312 2312
Continue to review full report at Codecov.
|
|
Will throw in some more comments later, but looks good so far! |
kylebystrom
changed the title
|
This should be ready to go now. |
| Parse the DFT input in a directory. | ||
| :param vasp_input: directory of vasp input | ||
| """ | ||
| return Structure.from_pmg_structure(Poscar.from_file(poscar).structure) |
There was a problem hiding this comment.
Can you ensure that this returns the cartesian coordinates every time since that's the FLARE convention?
There was a problem hiding this comment.
This is addressed by the unit test for the from_pmg_structure function, but I also will address it in the test_structure_parsing function when I assert the correct coordinate parsing
There was a problem hiding this comment.
Actually, test_struc.test_from_pmg_structure doesn't guarantee the cartesian coordinates are read because it sets the unit cell to np.eye(3). So I copy-pasted the test and changed the lattice parameters so that it will fail if from_pmg_structure returns fractional coordinates.
|
This looks great, Kyle! Can you address the two points above: 1. a small addition to the unit test, and 2. A guarantee that wherever in the code a FLARE structure is produced that the cartesian coordinates are returned? |
|
Closes #87 |
…ure to ensure Cartesian coordinates are read in
|
Okay, this is good to go now. Merging with master. |
The new vasp_util.py file can edit VASP POSCAR files with new coordinates, read forces from a static calculation, and read trajectories from an ionic relaxation/MD calculation. This is a draft; I still need to write unit tests and documentation.