Fix and cover RMSD by jdetle · Pull Request #889 · MDAnalysis/mdanalysis

jdetle · 2016-07-06T18:39:05Z

Fixes: #897

Changes made in this Pull Request:

Fixed a bug where we are still calling _process_selection as opposed to process_selection in RMSD (one that I think I created a while back)
- Changed RMSD to save numpy array.
- Added rmsdArray testfile
- Added RMSD tests

TODO

Currently left for another PR

Cover process_selection errors
Figure out how to force an overflow underflow error in RMSF

Currently stuck

Can open file handles be closed, if so, how?
Should tests always assert something? For example, I have a test to cover saving and weighting, but I don't actually check if the result is equal to some expected value.

coveralls · 2016-07-06T19:06:34Z

Coverage remained the same at 80.38% when pulling 18a4e4c on jdetle:process_selection into 7220db5 on MDAnalysis:develop.

coveralls · 2016-07-06T23:17:46Z

Coverage remained the same at 80.364% when pulling c868ed2 on jdetle:process_selection into 852daec on MDAnalysis:develop.

kain88-de · 2016-07-07T08:44:54Z

why wasn't this raising an error in the tests? Isn't this covered in the test?

jdetle · 2016-07-07T19:46:53Z

This coverage report seems to indicate not! . I will go ahead and write some tests for the RMSD class.

jdetle · 2016-07-07T19:49:05Z

It seems like RMSD is ripe for a refactor to the Bauhaus API too, so I could just go ahead and do that as well.

kain88-de · 2016-07-07T20:08:27Z

Yes if you like. Please finish the dmap first though

orbeckst · 2016-07-08T21:32:07Z

Leave refactor for later (as Max said). Fixing the bug and some tests are good enough for this PR.

Re: naming: I don't like RMSDSingleFrame, it still looks like a class. But ultimately your choice, just do it :-).

jdetle · 2016-07-08T21:33:08Z

On it.

coveralls · 2016-07-08T23:33:46Z

Coverage increased (+0.4%) to 80.774% when pulling f051b47 on jdetle:process_selection into 5d78f03 on MDAnalysis:develop.

jdetle · 2016-07-09T00:06:16Z

Now is as good as a time as any, what are open file handles and how do I close them?

coveralls · 2016-07-09T00:11:23Z

Coverage increased (+0.4%) to 80.805% when pulling 8b3d00a on jdetle:process_selection into 5d78f03 on MDAnalysis:develop.

orbeckst · 2016-07-09T01:01:54Z

    next (Issue #869)
  * Display of Deprecation warnings doesn't affect other modules anymore (Issue #754)
  * Changed nframes to n_frames in analysis modules for consistency (Issue #890)
+  * Removed call to protected process_selection function in RMSD


Leave this out from the CHANGELOG. I don't think that this was broeken in 0.15.0 or was it? If it was broken then this would be an entry under fixes, just stating that rms.RMSD was fixed (and reference #897)

Good point! Thanks.

coveralls · 2016-07-09T01:02:58Z

Coverage increased (+0.5%) to 80.897% when pulling 0358b44 on jdetle:process_selection into 5d78f03 on MDAnalysis:develop.

orbeckst · 2016-07-09T01:03:36Z

            if self.rmsd is None:
                raise NoDataError("rmsd has not been calculated yet")
-            np.savetxt(filename, self.rmsd)
+            np.save(filename, self.rmsd)


Why did you change this? This can break code for users, don't do this light-heartedly! Please revert!

save and savetxt are two very different things.

jdetle · 2016-07-09T01:14:24Z

There are a few mistakes and additional tests that I will go ahead and work on tonight. Beyond that, packing for a weekend in the bay before _SCIPY 2016_

coveralls · 2016-07-09T06:39:37Z

Coverage increased (+0.6%) to 81.013% when pulling 023408d on jdetle:process_selection into 5d78f03 on MDAnalysis:develop.

jdetle · 2016-07-09T06:59:37Z

Spurious test failures

orbeckst · 2016-07-09T17:14:58Z

Rerun the test on Travis. There should be a little reload button in the upper right. At least there is one for me when I have to do this.

Oliver Beckstein
email: orbeckst@gmail.com

Am Jul 8, 2016 um 23:59 schrieb John Detlefs notifications@github.com:

Spurious test failures

—
You are receiving this because you were assigned.
Reply to this email directly, view it on GitHub, or mute the thread.

jdetle · 2016-07-09T17:16:32Z

I have it but for some reason it throws an error saying the job can't be restarted, it's either a bug or a permissions issue.

jdetle · 2016-07-09T17:43:18Z

I will be away from the computer until late tonight. I hope everything looks alright. 1% coverage increase has to be nice :)

orbeckst · 2016-07-09T22:00:22Z

Travis appears to be happy now. Everything passed.

orbeckst · 2016-07-09T22:01:34Z

Still WIP or ready to be merged?

jdetle · 2016-07-09T22:03:35Z

Ready to be merged. All the todos are todone. There are still some minor
questions that I have about testing but merge-able nonetheless.
On Sat, Jul 9, 2016 at 3:01 PM Oliver Beckstein notifications@github.com
wrote:

Still WIP or ready to be merged?

—
You are receiving this because you authored the thread.
Reply to this email directly, view it on GitHub
#889 (comment),
or mute the thread
https://github.com/notifications/unsubscribe/AKcARg7JHi8iW2fNW77Hlq6usB7RvXrVks5qUBo_gaJpZM4JGYJ2
.

Have a wonderful day,
John J. Detlefs

orbeckst · 2016-07-09T22:05:20Z

Ah, just read the top post where you said:

Can open file handles be closed, if so, how?
Should tests always assert something? For example, I have a test to cover saving and weighting, but I don't actually check if the result is equal to some expected value

Answers:

Good question, @jbarnoud 's plugin reports files that are still open but I also don't have a clue how to fix these reported open files.
Each test should always assert something. At a minimum, check that a file was written, but better (and that's what we do in the trajectory writer tests), read the file back in and compare to the data that you used to write the file. If that is not feasible, manually read the file, compute some value (eg a mean) and check that you get it back from the saved file in the test.

orbeckst · 2016-07-09T22:07:33Z

   import matplotlib.pyplot as plt
   rmsd = R.rmsd.T   # transpose makes it easier for plotting
-   time = rmsd[1]
+   time = rmsd[1]facebook.com


There's a random "facebook.com" in the docs. Please remove.

orbeckst · 2016-07-09T22:18:31Z

@jdetle very good job! I'd like you to attend to the comments – primarily always assert something.

Thanks a lot for fixing and testing RMSD.

jdetle · 2016-07-10T04:25:45Z

Comments attended to!

coveralls · 2016-07-10T04:44:40Z

Coverage increased (+0.6%) to 81.015% when pulling 91ba82a on jdetle:process_selection into 5d78f03 on MDAnalysis:develop.

orbeckst · 2016-07-10T17:46:22Z

Thanks a lot @jdetle, good work!

Fixed (temporarily) broken analysis.rms.RMSD and added missing tests - fixes MDAnalysis#897 - full testing of the RMSD class Updated CHANGELOG * Added tests for RMSD, added a testfile. Changed RMSD class to save rmsd array as numpy file instead of text. * Tested saving * Test group selections Reverted back to save. Updated test coverage for exceptions. Found some errors with logging requiring a message. Reverted inclusion of unnecessary RMSD array. Fixed a small error and combined save and mass_weighted test. No Data Error. Covered RMSD.save(). * Fixed error messages * Asserted everywhere, fixed error messages.

jdetle force-pushed the process_selection branch from 18a4e4c to c868ed2 Compare July 6, 2016 22:54

jdetle mentioned this pull request Jul 7, 2016

RMSD, RMSF Refactor/Test Coverage #893

Closed

jdetle changed the title ~~Removed call to protected process_selection now that it isn't protected~~ [WIP] Fix RMSD, (which I broke) and add tests Jul 8, 2016

jdetle changed the title ~~[WIP] Fix RMSD, (which I broke) and add tests~~ [WIP] Fix and cover RMSD, (which I broke) Jul 8, 2016

jdetle changed the title ~~[WIP] Fix and cover RMSD, (which I broke)~~ [WIP] Fix and cover RMSD Jul 8, 2016

jdetle force-pushed the process_selection branch from 1ea25b1 to 856b56b Compare July 8, 2016 22:10

orbeckst added defect Component-Analysis Work in progress labels Jul 8, 2016

orbeckst added this to the 0.16.0 milestone Jul 8, 2016

orbeckst self-assigned this Jul 8, 2016

jdetle force-pushed the process_selection branch from 856b56b to f051b47 Compare July 8, 2016 23:09

jdetle changed the title ~~[WIP] Fix and cover RMSD~~ Fix and cover RMSD Jul 9, 2016

jdetle changed the title ~~Fix and cover RMSD~~ [WIP] Fix and cover RMSD Jul 9, 2016

orbeckst reviewed Jul 9, 2016
View reviewed changes

jdetle force-pushed the process_selection branch from e50390c to 023408d Compare July 9, 2016 06:15

jdetle changed the title ~~[WIP] Fix and cover RMSD~~ Fix and cover RMSD Jul 9, 2016

orbeckst removed the Work in progress label Jul 9, 2016

orbeckst reviewed Jul 9, 2016
View reviewed changes

jdetle added 2 commits July 9, 2016 20:41

Fixed error messages

40eb01c

Asserted everywhere, fixed error messages.

91ba82a

orbeckst merged commit e801d35 into MDAnalysis:develop Jul 10, 2016

jdetle mentioned this pull request Jul 11, 2016

Diffusion Maps rebase, new PR for final review #894

Merged

13 tasks

Uh oh!

Conversation

jdetle commented Jul 6, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

TODO

Currently left for another PR

Currently stuck

Uh oh!

coveralls commented Jul 6, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

coveralls commented Jul 6, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

kain88-de commented Jul 7, 2016

Uh oh!

jdetle commented Jul 7, 2016

Uh oh!

jdetle commented Jul 7, 2016

Uh oh!

kain88-de commented Jul 7, 2016

Uh oh!

orbeckst commented Jul 8, 2016

Uh oh!

jdetle commented Jul 8, 2016

Uh oh!

coveralls commented Jul 8, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jdetle commented Jul 9, 2016

Uh oh!

coveralls commented Jul 9, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

orbeckst Jul 9, 2016

Choose a reason for hiding this comment

Uh oh!

jdetle Jul 9, 2016

Choose a reason for hiding this comment

Uh oh!

coveralls commented Jul 9, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

orbeckst Jul 9, 2016

Choose a reason for hiding this comment

Uh oh!

jdetle commented Jul 9, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

coveralls commented Jul 9, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

jdetle commented Jul 9, 2016

Uh oh!

orbeckst commented Jul 9, 2016

Uh oh!

jdetle commented Jul 9, 2016

Uh oh!

jdetle commented Jul 9, 2016

Uh oh!

orbeckst commented Jul 9, 2016

Uh oh!

orbeckst commented Jul 9, 2016

Uh oh!

jdetle commented Jul 9, 2016

Uh oh!

orbeckst commented Jul 9, 2016

Uh oh!

orbeckst Jul 9, 2016

Choose a reason for hiding this comment

Uh oh!

jdetle Jul 10, 2016

Choose a reason for hiding this comment

Uh oh!

orbeckst commented Jul 9, 2016

Uh oh!

jdetle commented Jul 10, 2016

Uh oh!

coveralls commented Jul 10, 2016 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

jdetle commented Jul 6, 2016 •

edited

Loading

coveralls commented Jul 6, 2016 •

edited

Loading

coveralls commented Jul 6, 2016 •

edited

Loading

coveralls commented Jul 8, 2016 •

edited

Loading

coveralls commented Jul 9, 2016 •

edited

Loading

coveralls commented Jul 9, 2016 •

edited

Loading

jdetle commented Jul 9, 2016 •

edited

Loading

coveralls commented Jul 9, 2016 •

edited

Loading

coveralls commented Jul 10, 2016 •

edited

Loading