update helanal to AnalysisBase

[lilyminium]

Expected behavior

When I run helanal.helanal_trajectory, it produces readable output.

Actual behavior

helanal.helanal_trajectory writes 8 files (there is no way to turn this off):

• bending_matrix.dat
• fit_tilt.xvg
• local_bend.xvg
• local_tilt.xvg
• origin.pdb
• screw.xvg
• summary.txt
• unit_twist.xvg

None of the .xvg files have delimiters, resulting in something like this (unit_twist.xvg):

0173.50089152.36047157.11974168.0601134.1332682.78972102.89294480.54708134.09038143.31886166.2364105.43222124.6999295.75029122.1871694.43125101.5204499.2620497.0216893.1698991.6502786.83874136.04012134.16911119.11755132.66452166.35683171.27223122.5070494.534966102.8902595.2906597.6656998.8994297.9797.82383102.0121695.7262297.1709147.01093149.21326110.0322688.56357102.8416599.71714105.5752799.8775199.056145101.6334296.66122494.9870894.4209499.65817135.18112141.92354114.745056120.963615134.35696130.1924107.3315199.60976101.5804996.4396102.3210396.43453102.259056104.7718392.5648796.93780596.2955106.55468119.2636123.093636104.8041102.03329133.9388796.147766111.26341105.42476127.773766158.32767162.8811374.15202146.2733362.067436170.5584664.37851110.36818101.4832895.5074697.5645998.09011101.22569103.59741100.17954100.5833197.70348126.24198155.93248169.31538163.89807153.8199364.76133159.19398172.53307168.54634154.64725157.2409160.78711117.229416130.50992101.69997498.69098102.3421398.5284197.9638897.9031384.2355692.51822101.50628112.175995141.17561179.15326115.91146162.714297.18241148.72191177.6465675.098206162.01685178.84647153.61559143.4729582.728778.1872974.47471169.2817782.82487176.68405159.40784127.06898576.81265108.926895116.31709143.0244387.8953476.859406138.5434143.4651297.01566138.7108172.56514115.02348133.1791475.7148175.10802104.24445131.14348144.5514105.2866192.4977393.9222296.55736101.0048695.58964100.96390594.60849102.6927897.4528593.37078100.4637390.25625692.76536109.01559498.313156104.478905100.8604699.90523101.04384695.73096101.6445100.1566798.6450398.35263101.63966104.76728139.75473134.5733364.02203109.86525159.32852177.16191148.17155163.89886151.0398174.66324114.77936119.296104104.45007126.39088499.2082396.272104.2805799.6220799.19309105.1044101.61230598.236244101.4117290.6837682.89361

These aren't fixed-length either. Throwing in a newline after each value gives values with a mix of 4 or 5 decimal places.

0
173.50089
152.36047
157.11974
168.0601
134.13326
82.78972
102.892944
80.54708
134.09038
...

Am I missing some obvious way to interpret these files? If not, IMO these should either be written in an easy way for np.loadtxt to bring them back again, or left as attributes in a class.

The documentation also could use a lot of clarification and updating.

Code to reproduce the behavior

Show us how to reproduce the failiure. If you can, use trajectory files from the test data.

import MDAnalysis as mda
from MDAnalysis.tests.datafiles import PDB
from MDAnalysis.align import helanal
u = mda.Universe(PDB)
helanal.helanal_trajectory(u, 'name CA')

Currently version of MDAnalysis

• Which version are you using? (run python -c "import MDAnalysis as mda; print(mda.__version__)") '0.20.2-dev0'
• Which version of Python (python -V)? 3.7
• Which operating system? MacOS

[richardjgowers]

Iirc helanal might even predate numpy. I think rewriting the functionality into how analysis classes are meant to work (and saving against a class attribute) is probably a very good idea!

[IAlibay]

Just copying @wg150, I believe in our chat with @richardjgowers back in November he expressed having either started to work on this or wanting to. He was kinda busy with thesis work though...

[glass-w]

I'd be happy to take a look at this. I've already worked on changing the functionality and output formats. As @IAlibay has said, I'll most likely be able to work on this properly in the next few months.

[orbeckst]

A rewrite with AnalysisBase in mind would be most welcome!

Having a prototype sooner than later would be good so that we could decide if the new helanal should become the standard for release 1.0.

[orbeckst]

Rename this issue to "update helanal to AnalysisBase"?

[IAlibay]

Pinging @wg150, checking if you were still interested in picking this one up?

[lilyminium]

I would actually like to look at some helices sometime this week, so I’m also happy to poke at this if you’re busy @wg150

[glass-w]

I'll be able to take a look at this later this month. @lilyminium if you want to take a look / make a start this week that's ok with me!