Add version number function and version to bathy struct

[jstanleyx]

Add version 1.1 to existing master, which is version 1.1

[KateBrodie]

@jstanleyx I am testing this now and will approve when I confirm it works! looks good so far. -Kate

[KateBrodie]

@jstanleyx @RobHolman @mpalmsten

As I said above, I'm assuming the differences below are either (1) within the repeatability of cBathy or (2) the version currently saved at the FRF is from a slightly different version of the code. As a note, I also did this for a more recent file saved locally at the FRF and there were major differences along the seam, suggesting that the realtime data at the FRF is being run with what we will call version 1.2.

Below is the difference map (bathyOLD.fCombined.h - bathy.fCombined.h) to go with my comment above:

This was generated with the following code:

load('\\peeler\secure_share\argus\argus02b\2015\cx\251_Sep.08\1441729740.Tue.Sep.08_16_29_00.GMT.2015.argus02b.cx.cBathy.mat')
load('\\peeler\secure_share\argus\argus02b\2015\cx\251_Sep.08\1441729740.Tue.Sep.08_16_29_00.GMT.2015.argus02b.cx.mBW.mat')
bathyOLD = bathy;
bathy = analyzeBathyCollect(XYZ,T,RAW,bathyOLD)
figure;pcolor(bathyOLD.fCombined.h-bathy.fCombined.h);shading flat
caxis([-0.5 0.5])

[RobHolman]

Looks like seam problem differences. Will talk to John.

Rob Holman SECNAV/CNO Chair in Oceanography

104 Ocean Admin Bldg. CEOAS-OSU Corvallis, Oregon, USA 97331-5503 ph: 1-541-737-2914 holman@coas.oregonstate.edu http://cil-www.coas.oregonstate.edu

On Dec 11, 2017, at 2:30 PM, Katherine Brodie notifications@github.com wrote:

@jstanleyx @RobHolman @mpalmsten

Below is the difference map (bathyOLD.fCombined.h - bathy.fCombined.h) to go with my comment above:

This was generated with the following code:

load('\peeler\secure_share\argus\argus02b\2015\cx\251_Sep.08\1441729740.Tue.Sep.08_16_29_00.GMT.2015.argus02b.cx.cBathy.mat') load('\peeler\secure_share\argus\argus02b\2015\cx\251_Sep.08\1441729740.Tue.Sep.08_16_29_00.GMT.2015.argus02b.cx.mBW.mat') bathy = analyzeBathyCollect(XYZ,T,RAW,bathyOLD) figure;pcolor(bathyOLD.fCombined.h-bathy.fCombined.h);shading flat caxis([-0.5 0.5])

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

[KateBrodie]

@RobHolman maybe - as a contrast, when I ran the same code on 1509208140.Sat.Oct.28_16_29_00.GMT.2017 (the more recent file I spoke of above), the differences were definitely along the seam (see image below) - so I'm not sure the differences on the 2015 data are seam related...

[jstanleyx]

Version 1.1 has the seam problem. Master is currently version 1.1.

Pull request #28 into master from version-1.1 does nothing but add the version numbering.

Once this is approved and merged, "master" should be tagged as something to indicate the version. V1.1 or V1.1.0.

There is now a branch "version-1.2" which implements the seam fix and new initialization for better starting guesses. It also updates the version number.

Pull request #29 into "master" from "version-1.2" puts this all in master.

Testing should be:

run a file through master and through version-1.1. If the output is the same except for the "version" field in the bathy struct, then it should pass. The code is identical according to git, except for one function and one line in analyzeBathyCollect.m

I would approve it, except as "puller" I cannot.

then test the new master against version-1.2. The seam problem should be better, and the initial guesses should be better. I don't know how to define "better", that's a Rob-level thing.

Then tag the new master as V1.2.

The last step is to destroy the "development" branch and recreate it from master. Then development can proceed against that branch.

I do not think we should destroy the branches version-1.1 and version-1.2. They can serve as easy ways to get those versions.

On Tue, 2017-12-12 at 10:22 -0800, Rob Holman wrote:

Looks like seam problem differences. Will talk to John.

Rob Holman SECNAV/CNO Chair in Oceanography

104 Ocean Admin Bldg. CEOAS-OSU Corvallis, Oregon, USA 97331-5503 ph: 1-541-737-2914 holman@coas.oregonstate.edu http://cil-www.coas.oregonstate.edu

On Dec 11, 2017, at 2:30 PM, Katherine Brodie notifications@github.com wrote:

@jstanleyx @RobHolman @mpalmsten

Below is the difference map (bathyOLD.fCombined.h - bathy.fCombined.h) to go with my comment above:

This was generated with the following code:

load('\peeler\secure_share\argus\argus02b\2015\cx\251_Sep.08 \1441729740.Tue.Sep.08_16_29_00.GMT.2015.argus02b.cx.cBathy.mat') load('\peeler\secure_share\argus\argus02b\2015\cx\251_Sep.08 \1441729740.Tue.Sep.08_16_29_00.GMT.2015.argus02b.cx.mBW.mat') bathy = analyzeBathyCollect(XYZ,T,RAW,bathyOLD) figure;pcolor(bathyOLD.fCombined.h-bathy.fCombined.h);shading flat caxis([-0.5 0.5])

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

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

[KateBrodie]

@jstanleyx I understand the order of what we need to do - I was just also trying to check version 1.1 against the files that we have here at the FRF that we thought were run with version 1.1 (i.e. verify what's saved at the FRF). It's close but slightly different (first image). My test also showed that the newer data (2017) are clearly being run with something that has the seam fix (most likely the will be version 1.2) (see second image).

I did just run the same file through the master and version1.1 branches and the output matched, so I will approve.

[jstanleyx]

Ok. Sorry.

What you have from here from October of a couple of years ago (I think that's what you have) was run on a previous version of cBathy. I don't know what was in that one and it would take some serious investigation to figure it out.

More recent output is from version 1.2 that I have in production, but it does not yet have the version number stuff in it. That change was around the end of October.

On Tue, 2017-12-12 at 22:31 +0000, Katherine Brodie wrote:

@jstanleyx I understand the order of what we need to do - I was just also trying to check version 1.1 against the files that we have here at the FRF that we thought were run with version 1.1 (i.e. verify what's saved at the FRF). It's close but slightly different (first image). My test also showed that the newer data (2017) are clearly being run with something that has the seam fix (most likely the will be version 1.2) (see second image).

I did just run the same file through the master and version1.1 branches and the output matched, so I will approve.

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