Test function failed

[Xinqiu Yao]

We still get errors from R CMD check. Using auto_test() we found following messages. Can we get rid of them?

1. Failure(@test-nma.R#23): NMA ------------------------------------------------
nowU8 not equal to U8
Mean relative difference: 2

2. Failure(@test-nma.R#121): NMA -----------------------------------------------
nowMmtk7 not equal to mmtk7
Mean relative difference: 2

3. Failure(@test-nma.R#127): NMA -----------------------------------------------
nowMmtk7 not equal to mmtk7
Mean relative difference: 2

4. Failure(@test-nma.R#146): NMA -----------------------------------------------
nowMmtk7 not equal to mmtk7
Mean relative difference: 2

5. Failure(@test-nma.R#152): NMA -----------------------------------------------
nowMmtk7 not equal to mmtk7
Mean relative difference: 2

6. Failure(@test-nma.R#171): NMA -----------------------------------------------
nowAnm7 not equal to anm7
Mean relative difference: 2

Error in digest(path, file = TRUE) : 
  The specified pathname is not a file: /u2/xqyao/Project/bio3d/ver_devel/bio3d/data

[Lars Skjærven]

Hmm. what branch are you at?

> test_file("/home/slars/workspace/bio3d/ver_devel/bio3d/tests/testthat/test-nma.R")
Testing nma() : ..............................

[Xinqiu Yao]

I am on the latest master, rebuild bio3d and reinstall it. I tried the command you show above and it gives the same errors...

[Lars Skjærven]

hmm.. no errors on my linux box, but I do get errors on my mac. only the signs have changed:

> mmtk7
 [1]  9.399499e-03  9.162217e-03 -1.894026e-04  8.013488e-03  9.521462e-03
 [6]  3.004101e-04  6.725323e-03  6.130758e-03  7.167801e-04  3.911230e-03
[11]  3.103640e-03  1.122473e-04  5.015756e-05  1.223079e-03  5.064454e-04
[16]  1.487601e-03 -3.968054e-03  2.038939e-04
> nowMmtk7
 [1] -9.399497e-03 -9.162216e-03  1.894025e-04 -8.013487e-03 -9.521461e-03
 [6] -3.004100e-04 -6.725322e-03 -6.130757e-03 -7.167800e-04 -3.911229e-03
[11] -3.103640e-03 -1.122473e-04 -5.015756e-05 -1.223079e-03 -5.064454e-04
[16] -1.487601e-03  3.968053e-03 -2.038938e-04

With last R update I noticed the same change. Somehow this originates from the eigen() routine, but I haven't figured out why.

[Lars Skjærven]

Mac:

> sessionInfo()
R version 3.1.1 (2014-07-10)
Platform: x86_64-apple-darwin13.1.0 (64-bit)

Linux:

> sessionInfo()
R version 3.1.0 (2014-04-10)
Platform: x86_64-redhat-linux-gnu (64-bit)

[Barry Grant]

The signs here will be arbitrary and likely to change without warning on different builds. We have encountered this before with PCA routines also. There are old posts online about this, and new ones every so often.

[Xinqiu Yao]

If it is just sign changes, probably we can only compare the absolute values in test functions. What do you think?

[Xinqiu Yao]

Resolved with updated test-nma.R and test-nma.pdbs.R