[MITgcm-support] tutorial_global_oce_optim optimisation failed

Andrew McRae andrew.mcrae at physics.ox.ac.uk
Tue Jun 19 10:37:46 EDT 2018 

Okay, I have
1) copied OPENAD_OPTIONS.h from pkg/openad to the code_oad/ subfolder of
the tutorial, changing it to define ALLOW_OPENAD_ACTIVE_READ_XY

Good news: the main body of tools/OAD_support/ad_template.active_read_xy.F
(which is wrapped in #ifdef ALLOW_OPENAD_ACTIVE_READ_XY) now appears in
external_Dummies_cb2m_oad.f

Bad news: this gives a compile error in externalDummies_cb2m_oad.f of
"Error: Unexpected '%' for nonderived-type variable 'active_var'".  This
seems to be because active_var is declared as a REAL(w2f__8) in
externalDummies_cb2m_oad.f, not a type(active).  The lines of code
corresponding to

      active_var = dummy + active_var
      dummy = active_var(1,1,1,1) + dummy

don't appear in the post-processed code [optimized out by the OpenAD
toolchain, or something else?], which is probably why active_var doesn't
become an active variable.  Therefore, I....

2) change the type of active_var to type(active) in the post-processed file
(yuck).  make adAll continues from where it left off, and mitgcmuv_ad now
compiles :)

(I tried changing the type of this variable in pkg/openad/externalDummies.F
<https://github.com/MITgcm/MITgcm/blob/master/pkg/openad/externalDummies.F#L285>,
but this leads to a bork in the OpenAD toolchain)

I can confirm the cost function changes from iteration to iteration, and
I'll now test if the optimization works.  Hopefully you can find a more
permanent solution to the above.

Andrew

On 19 June 2018 at 13:43, Andrew McRae <andrew.mcrae at physics.ox.ac.uk>
wrote:

> The active_read_xy routine used in OpenAD mode looks suspicious:
> https://github.com/MITgcm/MITgcm/blob/master/pkg/openad/exte
> rnalDummies.F#L269-L296
>
> 1) ALLOW_OPENAD_ACTIVE_READ_XY isn't defined for
> tutorial_global_oce_optim; I guess it should be?
>
> 2) This routine seems to be basically a no-op anyway?  I guess
> active_var_file should be read into active_var, or similar?
>
> Andrew
>
> On 18 June 2018 at 18:04, Andrew McRae <andrew.mcrae at physics.ox.ac.uk>
> wrote:
>
>> Not sure if you've had a chance to look at this yet... the only time I
>> can see tmpfld2d being written to (and not just initialised to 0.0 or 1.0)
>> is in pkg/admtlm/bypassad.F line 96.  Presumably that package isn't
>> switched on here.  I can't see xx_hfluxm being written to at all.
>>
>> A few lines above, active_read_xy is called with xx_hfluxm_dummy as the
>> last argument... should this have been xx_hfluxm, perhaps?
>> (xx_hfluxm_dummy is a single variable, while xx_hfluxm is an array, so this
>> probably won't work as-is...)
>>
>> Andrew
>>
>> On 13 June 2018 at 23:18, Andrew McRae <andrew.mcrae at physics.ox.ac.uk>
>> wrote:
>>
>>> Okay, thank you.  If do you have any advice on debugging this, do say.
>>> I guess you already got as far as spotting that all the terms on the RHS of
>>> https://github.com/MITgcm/MITgcm/blob/master/pkg/ctrl/ctrl_m
>>> ap_forcing.F#L259 are zero.
>>>
>>> Andrew
>>>
>>> On 13 June 2018 at 21:36, Patrick Heimbach <heimbach at mit.edu> wrote:
>>>
>>>> Andrew,
>>>>
>>>> I have not been able to look into this due to various other commitments
>>>> over the last couple of months.
>>>>
>>>> I'll be grounded for a while in Austin starting next week, and this
>>>> will be near the top of my ToDo list.
>>>>
>>>> Patrick
>>>>
>>>> > On Jun 13, 2018, at 12:56 PM, Andrew McRae <
>>>> andrew.mcrae at physics.ox.ac.uk> wrote:
>>>> >
>>>> > MITgcm built with OpenAD is not making use of the ecco_ctrl files for
>>>> optimcycle >= 1.  The file apparently gets read in, but the contents get
>>>> dropped on the floor somewhere.
>>>> >
>>>> > Andrew
>>>> >
>>>> > On 13 June 2018 at 18:51, Matthew Mazloff <mmazloff at ucsd.edu> wrote:
>>>> > Hello
>>>> >
>>>> > Sorry, I lost track. What needs to be debugged? Can you please
>>>> reiterate the problem?
>>>> >
>>>> > Thanks
>>>> > Matt
>>>> >
>>>> >
>>>> >> On Jun 13, 2018, at 10:14 AM, Andrew McRae <
>>>> andrew.mcrae at physics.ox.ac.uk> wrote:
>>>> >>
>>>> >> Hi Patrick,
>>>> >>
>>>> >> Were you able to make any progress with this?  If not, do you have
>>>> any advice on debugging this?  (I'm getting lost in ctrl_unpack as to which
>>>> variable the control vector is even read into)
>>>> >>
>>>> >> Thanks,
>>>> >> Andrew
>>>> >>
>>>> >> On 5 May 2018 at 20:12, Patrick Heimbach <heimbach at mit.edu> wrote:
>>>> >> A quick update:
>>>> >>
>>>> >> This tutorial works as advertised (in the manual), but not as
>>>> "hoped".
>>>> >> What I mean is that it has been developed and only ever fully tested
>>>> and used  in optimization mode with TAF-generated code (and that's what's
>>>> documented in the manual).
>>>> >>
>>>> >> Of course, it should not make a difference of whether we use TAF vs.
>>>> OpenAD as long as gradients are correct. But as it turns out, with the
>>>> OpenAD code there appears to be a little glitch. Gradient seems correct,
>>>> and iteration 1 update is properly read in, but then not used (instead it
>>>> is reset to zero). Oh well. I'll need to check where that happens, so stay
>>>> tuned.
>>>> >>
>>>> >> p.
>>>> >>
>>>> >> > On May 4, 2018, at 10:11 AM, Andrew McRae <
>>>> andrew.mcrae at physics.ox.ac.uk> wrote:
>>>> >> >
>>>> >> > And, still no luck(?)
>>>> >> >
>>>> >> > Running for a year (switching the commented and uncommented
>>>> nTimeSteps and lastinterval declarations in data and data.cost), optim.x
>>>> (lsopt+optim, not optim_m1qn3) now gives the output
>>>> >> >
>>>> >> >   cost function............... 0.60514949E+01
>>>> >> >   norm of x................... 0.00000000E+00
>>>> >> >   norm of g................... 0.23235517E+00
>>>> >> >
>>>> >> >   optimization stopped because :
>>>> >> >   ifail =   4    the search direction is not a descent one
>>>> >> >
>>>> >> >
>>>> >> > On 4 May 2018 at 13:58, Andrew McRae <
>>>> andrew.mcrae at physics.ox.ac.uk> wrote:
>>>> >> > On 4 May 2018 at 06:04, Patrick Heimbach <heimbach at mit.edu> wrote:
>>>> >> > Hi Matt,
>>>> >> >
>>>> >> > as you indicated, all is still good, and I suspect the same you
>>>> did regarding what might be at issue.
>>>> >> >
>>>> >> > I just downloaded latest MITgcm, re-ran adjoint, and conducted 2
>>>> iterations (using lsopt).
>>>> >> >
>>>> >> > It still works "out of the box" ... if one realizes that a manual
>>>> is part of that "box", and section 3.18 (old manual prior to readthedocs)
>>>> has some description of this tutorial, thanks to dfer (admittedly somewhat
>>>> out of date, but still mostly relevant). In particular it says there that
>>>> the optimization has been conducted for a 1-year simulation.
>>>> >> >
>>>> >> > Okay, thanks.  I interpreted the manual footnote as "running a
>>>> 1-year simulation will reproduce the scientifically-interesting graphs in
>>>> the manual", not as "the default parameters are only useful for verifying
>>>> correctness of the adjoint, but will break the optimisation routine".  I'll
>>>> see if I have more success with the longer run.
>>>> >> >
>>>> >> >
>>>> >> > Since we do not want to conduct 1-year integrations for *any* of
>>>> the tutorials within our regression tests (these tests consist of 90
>>>> forward, 24 adjoint/TAF, 10 adjoint/OpenAD, and 16 tangent-linear/TAF
>>>> configurations, each needing to be compiled and executed) we have shortened
>>>> the number of time steps to 10 (= 10 days) to perform efficient nightly
>>>> regression tests of the adjoint. Not changing the number of time steps
>>>> leads to optimizing in the noise - in fact cost function goes up in that
>>>> case.
>>>> >> >
>>>> >> > That the user's cost function does not change at all suggests a
>>>> more basic problem though (hard to speculate what it might be).
>>>> >> >
>>>> >> > I made a quick test by extending nTimeSteps from 10 to 90 days,
>>>> which leads to cost reduction as desired, namely, for:
>>>> >> >  numiter=1,
>>>> >> >  nfunc=3,
>>>> >> >  fmin=5.74,
>>>> >> > (values in data.optim that comes with tutorial_global_oce_optim)
>>>> >> > I obtain following costs:
>>>> >> > iter. 0: fc =  0.184199260445164D+02
>>>> >> > iter. 1: fc =  0.130860446841901D+02
>>>> >> > iter. 2: fc =  0.979374136987667D+01
>>>> >> >
>>>> >> > I did that test "by hand", i.e. not using the script cycsh also
>>>> provided (see manual). Doing so by hand requires two more lines in
>>>> data.ctrl:
>>>> >> >  &CTRL_PACKNAMES
>>>> >> >  costname='ecco_cost',
>>>> >> >  ctrlname='ecco_ctrl',
>>>> >> >
>>>> >> > Since gradients produced with TAF are extremely similar (10+
>>>> digits?) to those produce with OpenAD (see results/ directory which has
>>>> both TAF and OpenAD reference results), I expect it to work with OpenAD too
>>>> (have not tested it right now).
>>>> >> >
>>>> >> > -Patrick
>>>> >> >
>>>> >> >
>>>> >> >
>>>> >> > > On May 2, 2018, at 12:34 PM, Andrew McRae <
>>>> andrew.mcrae at physics.ox.ac.uk> wrote:
>>>> >> > >
>>>> >> > > Thanks for this.
>>>> >> > >
>>>> >> > > Just as a sanity check, before I involve optim_m1qn3 again, the
>>>> output of my ./testreport -t tutorial_global_oce_optim -oad includes
>>>> >> > >
>>>> >> > > There were 16 decimal places of similarity for "ADM CostFct"
>>>> >> > > There were 16 decimal places of similarity for "ADM Ad Grad"
>>>> >> > > There were 0 decimal places of similarity for "ADM FD Grad"
>>>> >> > >
>>>> >> > > Should I be concerned about this?
>>>> >> > >
>>>> >> > > E.g. lines 2116-2118 of my output_oadm.txt file are
>>>> >> > >
>>>> >> > > (PID.TID 0000.0001)  ADM  ref_cost_function      =
>>>> 6.20023228182329E+00
>>>> >> > > (PID.TID 0000.0001)  ADM  adjoint_gradient       =
>>>> -2.69091500991183E-06
>>>> >> > > (PID.TID 0000.0001)  ADM  finite-diff_grad       =
>>>> 0.00000000000000E+00
>>>> >> > >
>>>> >> > > But at least my cost function value is the same:
>>>> >> > >
>>>> >> > > (PID.TID 0000.0001)   local fc =  0.620023228182329D+01
>>>> >> > > (PID.TID 0000.0001)  global fc =  0.620023228182329D+01
>>>> >> > >
>>>> >> > > Andrew
>>>> >> > >
>>>> >> > > On 2 May 2018 at 10:34, Martin Losch <Martin.Losch at awi.de>
>>>> wrote:
>>>> >> > > Hi Andrew,
>>>> >> > >
>>>> >> > > I won’t be able to help you much with the optim/lsopt code,
>>>> because I would have to get it running again myself. But I do recommend
>>>> using the MITgcm_contrib/mlosch/optim_m1qn3 code. It’s not very well
>>>> documented, but I am attaching a skeleton script to illustrate how to use
>>>> it. Please give it a try and if you find it useful, I can add this script
>>>> to the repository.
>>>> >> > >
>>>> >> > > The two versions of the optimization routine are similar, both
>>>> implement the same optimization algorithm (BFGS), but optim_m1qn3 uses a
>>>> later version of the m1qn3 code, I think it’s easier to compile (only one
>>>> Makefile) and I believe (but there’s debate about this) that it does the
>>>> right thing as opposed to the optim/lsopt variant, which somehow truncates
>>>> the optimization in each iteration. Having said that, I have used both in
>>>> parallel, and the reduction of the cost function (which is really all we
>>>> care about) is sometimes better with the optim_m1qn3 code, sometimes it is
>>>> better with the optim/lsopt code. The optim_m1qn3 code is closer to the
>>>> idea of the original m1qn3 code.
>>>> >> > >
>>>> >> > > Let me know if you can use my attached instructions.
>>>> >> > >
>>>> >> > > Martin
>>>> >> > >
>>>> >> > >
>>>> >> > >
>>>> >> > > > On 1. May 2018, at 00:00, Andrew McRae <
>>>> andrew.mcrae at physics.ox.ac.uk> wrote:
>>>> >> > > >
>>>> >> > > > Right, but the cost function is the same value each time, the
>>>> norm of x is 0 each time, and the norm of g is the same each time.  This
>>>> suggests nothing is happening.  It's a bit ridiculous that one of the core
>>>> tutorials simply isn't working out of the box...
>>>> >> > > >
>>>> >> > > > I will have a go at debugging.
>>>> >> > > >
>>>> >> > > > Andrew
>>>> >> > > >
>>>> >> > > > On 30 April 2018 at 22:54, Matthew Mazloff <mmazloff at ucsd.edu>
>>>> wrote:
>>>> >> > > > Well you are correct that its not actually taking a step
>>>> because the dot product of the control is 0:
>>>> >> > > >>> norm of x................... 0.00000000E+00
>>>> >> > > > meaning the controls are all 0 still.
>>>> >> > > >
>>>> >> > > > However the gradients are non-zero
>>>> >> > > >>> norm of g................... 0.12730927E-01
>>>> >> > > > so the linesearch should step and
>>>> >> > > > ecco_ctrl_MIT_CE_000.opt0001
>>>> >> > > > should not be all zero.
>>>> >> > > >
>>>> >> > > > To debug this you could put a print statement in
>>>> optim_writedata.F to see what it is writing…..
>>>> >> > > >
>>>> >> > > > I don’t know enough about this tutorial to be a bigger help,
>>>> sorry
>>>> >> > > >
>>>> >> > > > Matt
>>>> >> > > >
>>>> >> > > >
>>>> >> > > >> On Apr 30, 2018, at 2:50 PM, Andrew McRae <
>>>> andrew.mcrae at physics.ox.ac.uk> wrote:
>>>> >> > > >>
>>>> >> > > >> Yes, I did.
>>>> >> > > >>
>>>> >> > > >> On 30 April 2018 at 22:42, Matthew Mazloff <mmazloff at ucsd.edu>
>>>> wrote:
>>>> >> > > >> This is still iteration 0. You have to update data.optim to
>>>> tell it you are now at iteration 1
>>>> >> > > >>
>>>> >> > > >> Matt
>>>> >> > > >>
>>>> >> > > >>
>>>> >> > > >>> On Apr 30, 2018, at 2:38 PM, Andrew McRae <
>>>> andrew.mcrae at physics.ox.ac.uk> wrote:
>>>> >> > > >>>
>>>> >> > > >>> I tried a few steps of this, but the output of optim.x
>>>> always has
>>>> >> > > >>>
>>>> >> > > >>>   cost function............... 0.62002323E+01
>>>> >> > > >>>   norm of x................... 0.00000000E+00
>>>> >> > > >>>   norm of g................... 0.12730927E-01
>>>> >> > > >>>
>>>> >> > > >>> near the end, with no decrease in the cost function.  So I
>>>> guess it's not actually taking the step?
>>>> >> > > >>>
>>>> >> > > >>> Andrew
>>>> >> > > >>>
>>>> >> > > >>> On 27 April 2018 at 18:04, Andrew McRae <
>>>> andrew.mcrae at physics.ox.ac.uk> wrote:
>>>> >> > > >>> !!!  Okay...
>>>> >> > > >>>
>>>> >> > > >>> Yes, it produced the .opt0001 file.  I'll see how this goes.
>>>> >> > > >>>
>>>> >> > > >>> Thanks,
>>>> >> > > >>> Andrew
>>>> >> > > >>>
>>>> >> > > >>> On 27 April 2018 at 17:57, Matthew Mazloff <
>>>> mmazloff at ucsd.edu> wrote:
>>>> >> > > >>> Hello
>>>> >> > > >>>
>>>> >> > > >>> Its been awhile, but I am pretty sure that is the normal
>>>> output. It says “fail", but it did give you a new and
>>>> ecco_ctrl_MIT_CE_000.opt0001 (correct?) and if you unpack and run likely
>>>> the cost will descend.
>>>> >> > > >>>
>>>> >> > > >>> I think it worked correctly. lsopt/optim are just
>>>> confusing…but I think its working. I think all is good!
>>>> >> > > >>>
>>>> >> > > >>> Matt
>>>> >> > > >>>
>>>> >> > > >>>
>>>> >> > > >>>
>>>> >> > > >>>> On Apr 27, 2018, at 8:25 AM, Andrew McRae <
>>>> andrew.mcrae at physics.ox.ac.uk> wrote:
>>>> >> > > >>>>
>>>> >> > > >>>> Just separating this from the other thread, I got the
>>>> bundled MITgcm optim routine built (having made these changes, based on
>>>> this thread from 2010 and this one from 2016).
>>>> >> > > >>>>
>>>> >> > > >>>> I use OpenAD to create the adjoint.
>>>> >> > > >>>>
>>>> >> > > >>>> My steps are:
>>>> >> > > >>>> 1) in the build directory, run ../../../tools/genmake2 -oad
>>>> -mods=../code_oad
>>>> >> > > >>>> 2) run make depend and make adAll
>>>> >> > > >>>> 3) copy input_oad/ into a new folder scratch/
>>>> >> > > >>>> 4) within scratch/, run ./prepare_run
>>>> >> > > >>>> 5) copy mitgcmuv_ad from build/ into scratch/, copy optim.x
>>>> into scratch/OPTIM/
>>>> >> > > >>>> 6) run ./mitgcmuv_ad
>>>> >> > > >>>> 7) in scratch/OPTIM, create symlinks to ../data.optim and
>>>> ../data.ctrl
>>>> >> > > >>>> 8) copy the files ecco_cost_MIT_CE_000.opt0000 and
>>>> ecco_ctrl_MIT_CE_000.opt0000 into the OPTIM subdirectory
>>>> >> > > >>>> 9) run ./optim.x within the subdirectory
>>>> >> > > >>>>
>>>> >> > > >>>> The full output is attached, but I assume the optimisation
>>>> failed since the last lines are
>>>> >> > > >>>>
>>>> >> > > >>>>   optimization stopped because :
>>>> >> > > >>>>   ifail =   4    the search direction is not a descent one
>>>> >> > > >>>>
>>>> >> > > >>>> Any ideas?  (I guess this isn't something that is tested in
>>>> the daily builds?)
>>>> >> > > >>>>
>>>> >> > > >>>> In the meantime, I'll try the m1qn3 routine as in the other
>>>> thread, which should help distinguish between a problem with the
>>>> optimisation routine or the gradient generated by mitgcmuv_ad.
>>>> >> > > >>>>
>>>> >> > > >>>> Andrew
>>>> >> > > >>>> <out.txt>_______________________________________________
>>>> >> > > >>>> MITgcm-support mailing list
>>>> >> > > >>>> MITgcm-support at mitgcm.org
>>>> >> > > >>>> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>>>> >> > > >>>
>>>> >> > > >>>
>>>> >> > > >>>
>>>> >> > > >>> _______________________________________________
>>>> >> > > >>> MITgcm-support mailing list
>>>> >> > > >>> MITgcm-support at mitgcm.org
>>>> >> > > >>> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>>>> >> > > >>
>>>> >> > > >>
>>>> >> > > >> _______________________________________________
>>>> >> > > >> MITgcm-support mailing list
>>>> >> > > >> MITgcm-support at mitgcm.org
>>>> >> > > >> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>>>> >> > > >
>>>> >> > > >
>>>> >> > > > _______________________________________________
>>>> >> > > > MITgcm-support mailing list
>>>> >> > > > MITgcm-support at mitgcm.org
>>>> >> > > > http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>>>> >> > >
>>>> >> > > _______________________________________________
>>>> >> > > MITgcm-support mailing list
>>>> >> > > MITgcm-support at mitgcm.org
>>>> >> > > http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>>>> >> > >
>>>> >> > > _______________________________________________
>>>> >> > > MITgcm-support mailing list
>>>> >> > > MITgcm-support at mitgcm.org
>>>> >> > > http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>>>> >> >
>>>> >> >
>>>> >> >
>>>> >> > _______________________________________________
>>>> >> > MITgcm-support mailing list
>>>> >> > MITgcm-support at mitgcm.org
>>>> >> > http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>>>> >>
>>>> >>
>>>> >> _______________________________________________
>>>> >> MITgcm-support mailing list
>>>> >> MITgcm-support at mitgcm.org
>>>> >> http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>>>> >
>>>> >
>>>> > _______________________________________________
>>>> > MITgcm-support mailing list
>>>> > MITgcm-support at mitgcm.org
>>>> > http://mailman.mitgcm.org/mailman/listinfo/mitgcm-support
>>>>
>>>>
>>>
>>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.mitgcm.org/pipermail/mitgcm-support/attachments/20180619/1deff1cf/attachment-0001.html>

More information about the MITgcm-support mailing list