XYZ elements

[RMeli]

Fixes #2420 and #2421

Changes made in this Pull Request:

• XYZParser adds elements attribute
• XYZWriter tries to use the atoms.elements attribute (fallback to atoms.names)

PR Checklist

• Tests
• CHANGELOG updated
• Issue raised/referenced

[RMeli]

Here I assume that names contain element symbols (XYZ file format). As discusses in #2420 this might not always be the case. Should I make this more flexible?

[richardjgowers]

I think (hope) guess_masses is smart enough to skip a bad value and print a warning?

[RMeli]

@richardjgowers tests for the writer are still missing...

Any suggestions on how to best write them?

[richardjgowers]

@RMeli something like...

import io
import MDAnalysis as mda

u = mda.Universe.empty(n_atoms=5)

u.add_TopologyAttr('elements', values=['Te', 'S', 'Ti', 'N', 'Ga'])

ns = mda.lib.util.NamedStream(io.StringIO(), 'file.xyz')

with mda.Writer(ns) as w:
    w.write(u)

names = ''.join(l.split()[0] for l in ns.readlines())

assert names[:-1].lower() == 'testing'

[RMeli]

@richardjgowers Thanks for the funny suggestion. I meant to ask how to best incorporate them in the existing suite... I put the tests in TestXYZWriterNames and the they try writing with both elements and names attributes.

[codecov]

Codecov Report

Merging #2456 into develop will increase coverage by 0.01%.
The diff coverage is 83.33%.

@@             Coverage Diff             @@
##           develop    #2456      +/-   ##
===========================================
+ Coverage    89.99%   90.01%   +0.01%     
===========================================
  Files          177      177              
  Lines        22439    22233     -206     
  Branches      2914     2908       -6     
===========================================
- Hits         20193    20012     -181     
+ Misses        1637     1619      -18     
+ Partials       609      602       -7

Impacted Files	Coverage Δ
package/MDAnalysis/coordinates/XYZ.py	87.97% <ø> (-0.16%)	⬇️
package/MDAnalysis/topology/XYZParser.py	100% <ø> (ø)	⬆️
package/MDAnalysis/visualization/streamlines.py	88.18% <ø> (ø)	⬆️
package/MDAnalysis/analysis/waterdynamics.py	82.68% <ø> (+0.17%)	⬆️
package/MDAnalysis/visualization/streamlines_3D.py	94.08% <ø> (ø)	⬆️
package/MDAnalysis/coordinates/memory.py	96.05% <0%> (ø)	⬆️
package/MDAnalysis/topology/TXYZParser.py	100% <100%> (ø)	⬆️
package/MDAnalysis/coordinates/TXYZ.py	90.1% <100%> (-0.22%)	⬇️
package/MDAnalysis/lib/util.py	88.47% <100%> (+1.08%)	⬆️
package/MDAnalysis/auxiliary/base.py	86.11% <50%> (-0.86%)	⬇️
... and 49 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 11c5bab...3fca537. Read the comment docs.

[orbeckst]

Given that @richardjgowers approved and I checked the codecov is nothing to worry about (the low diff coverage comes from changes in indentations that don't even show up in these files – weird) so I am merging,