Dipole grid for the ions #160

abukowski21 · 2024-12-18T04:37:07Z

Dipole grid for the ions

Don't see an issue for this... But this PR modifies/overhauls the dipole grid for the ions.

Within edu/example/dipole/ there is a Python file which will generate
plots from this new grid. It is helpful to experiment with that to see how
the grid is affected by different parameters

The methodology in that file is pretty outdated, but the locations are
OK. If requested, I can update that file to use the same methods to
what's being put into src/init_mag_grid.cpp.

Going to go through a little explanation here, since the code is complicated.

Inputs are min_alt, min_apex, LatMax, LatStretch, dAltStretch, and of course, the number of cells in each direction
- min_alt defines the minimum altitude of the cell centers
- min_apex is the altitude of the lowest field line. This controls the minimum latitude.
- LatMax sets, roughly, the maximum latitude field lines are traced from. It's not exact because things are complicated.
  - When the user requests a single longitude dimension, field lines are traced from about halfway between LatMax and min_alt. I set the dLat to ~1 degree, same with longitude.
- LatStretch is not currently used. So it can only be set to one, otherwise the code will complain and change it to 1. LatStretch=1 corresponds to linear spacing of base latitudes and eventually this will control the spacing of the base latitudes.
- dAlt stretch controls how points are laid down along magnetic field lines. Values much less than one put more points at lower altitudes, and values closer to one put more points higher. I like the look of ~0.2.

I can document more and/or include some more plotting scripts that show what things are doing, if anyone wants.

Type of change

Please delete options that are not relevant.

Bug fix (non-breaking change that fixes an issue) kinda: there's no issue for this, but it is an "issue"
Breaking change (fix or feature that would cause existing functionality
to not work as expected)
This change requires a documentation update

How Has This Been Tested?

1D, 2D, 3D, with a large number of possible resolutions.
- They all work as expected
- There are some configurations that put corners on the pole. This does not (yet) raise errors. But those points just don't exist
I checked orthogonality, expected locations, conversions to geographic.... everything I can think of.
- They all work as I would expect now
- That's why this took so long
- Running on one core causes some weird things to happen, which I'll make issues for.

Test configuration

gcc 13.3.0 on WSL2 (Ubuntu) on a Windows machine. Ran on Linux too with gcc 13.2.0.

Checklist:

[N/A] Make sure you are merging into the develop (not master) branch
My code follows the style guidelines of this project
I have performed a self-review of my own code
I have commented my code, particularly in hard-to-understand areas
I have made corresponding changes to the documentation
My changes generate no new warnings
I have added tests that prove my fix is effective or that my feature works
New and existing unit tests pass locally with my changes
Any dependent changes have been merged and published in downstream modules
[N/A] Add a note to CHANGELOG.md, summarizing the changes

(It's not used yet)

4/3, not 3/3. taking (normalized) 2/3 step each time,

-> the ghost cells aren't in yet though. changes to grid.cpp will support any variation of the dipole grid, from dipole to dipole6. Any number, really.

rearrange where inputs are checked too.

latitudes are backwards now???

… (probably won't be used)

it's weird to do it this way, but it works

and a comment describing what's going on.

- clean up comments & report statements (typos) - lay down baselats in the same order as geo grid, make interpolation easier later!

- the order stuff was placed was weird. - correct comments too

- Pretty sure this is correct...

…ers later

forgot to commit this earlier. Scale the exp_q_dist by nAlts, not nAlts/2 We can also switch to q_Start (and 0), instead of q_s and q_N

in an effort to try and figure out why its wrong...

…_mag_grid

…l vars if a var within a function has the same name as a global & it changes the value, let's just be really clear it's only the **local** var that should be changed.

found a couple unused variables, and some that shared names w others

holy crap this was hard to find. Dipole looks (mostly) good now

> this is "bad github practice" but there were some changes that got forgotten in previous commits. whoops. it's all going to the same place anyways, as they say :)

This could break old input files, watch out! Seems ok on my end though...

…n mag_grid also move a couple functions around Just cleanin' up the code a bit

holy crap this has been a week of hunting the smallest bugs and chasing down the most miniscule of errors Dipole looks great now, for realz this time. Works in 1 & 3 dimensions This commit makes baselats_down a global var, and changes how the corners are put in. There are no longer any zeros, nans, etc. note, some configurations of the grid *can* result in nan's at the poles, but these are rare - code beforte this commit would have the last field line shifted one q-step lower than it should be. this makes everything even & makes sure the centers and corners are lined up.

whoops. aether doesn't need this lol

abukowski21 · 2024-12-20T00:49:06Z

Um just did a merge commit with the web editor. I hope it's right. If the checks pass, then this should be good to go!

abukowski21 added 30 commits October 3, 2024 19:55

[FEAT]: Create ion quadtree

f583540

(It's not used yet)

[BUG]: Correct origin location in sphere6 and dipole6

eaa8695

4/3, not 3/3. taking (normalized) 2/3 step each time,

[feat]: Move dipole grid funcs to their own header file

5271b5f

[bug] add support for quadtree into dipole grid.

8b937af

-> the ghost cells aren't in yet though. changes to grid.cpp will support any variation of the dipole grid, from dipole to dipole6. Any number, really.

[bug]: init_dipole_grid's errors are now handled correctly.

3d9472a

rearrange where inputs are checked too.

[bug]: Update default input files to use 1 for baselat exponent

40dd52d

FEAT: Dipole-> cartesian transform

b21d559

FEAT: Dipole grid spacing. Now in 3D!(irections)

d02467f

BUG: Wrong onject types in fill_grid inlines

d751750

FEAT: mag2geo, spacing, etc. for dipole grid

1048e80

feat: Store r,theta better for dipole grid

b6dea8c

Clean up spacing along field line

58dbcde

STY: Mostly docs, comments, debugging statements, etc.

87bee6f

update headers to include init_mag_griod func's

fcf5ab9

BUG: Couple more issues fixed.

652f92c

latitudes are backwards now???

STY: clean up comment whitespace

0f9f737

Clean up some of the debugging & create a template for dipole gravity…

e012e96

… (probably won't be used)

sty: clean up comments

839f723

BUG: Fix indexing when inserting cells into dipole grid.

e2d00c9

it's weird to do it this way, but it works

STY: Formatting fill_grid.cpp

87511ed

and a comment describing what's going on.

doc: update comments in init_mag_grid field line spacing

1ff1b9d

BUG/DOC: Fixes in baselat spacing in init_mag_grid

ee7e07f

- clean up comments & report statements (typos) - lay down baselats in the same order as geo grid, make interpolation easier later!

DOC/BUG: field-line spacing bug & docs in init_mag_grid

e78d6db

- the order stuff was placed was weird. - correct comments too

FEAT: Dipole edges: longitude

f4cd32a

FEAT Dipole edges: latitude

80bdf83

FEAT: Set sizes of unique dipole edge cubes

a842468

FEAT: (q,p) -> (r,theta) func takes cubes as input now

3ea99ef

- Pretty sure this is correct...

FEAT: Lay down points on field line so it's easier to get to the corn…

b59854b

…ers later

FEAT: Fix the dipole corners. Updates to all grid files

24406d9

BUG: Lay down a half, not whole, dipole field line

7455f68

forgot to commit this earlier. Scale the exp_q_dist by nAlts, not nAlts/2 We can also switch to q_Start (and 0), instead of q_s and q_N

abukowski21 added 18 commits December 18, 2024 14:03

STY: minor linting, some readability things in init_mag_grid & grid.h

6c1ae57

bug/feat Pass planet radius into dipole corner function

daf6cfe

in an effort to try and figure out why its wrong...

bug/sty: dont rely on data being initialized to zero for magP in init…

11ba535

…_mag_grid

sty (bug?): make init_mag_grid be SUPER pedantic about local vs globa…

142e5d7

…l vars if a var within a function has the same name as a global & it changes the value, let's just be really clear it's only the **local** var that should be changed.

sty: more clear naming of vars, again, in init_mag_grid

f98514e

found a couple unused variables, and some that shared names w others

BUG: Fix math in qp->r,theta for arma_cube in init_mag grid

cbce965

holy crap this was hard to find. Dipole looks (mostly) good now

BUGS: More robust handling of dipole corners, cleanup from last commits

35e35ed

> this is "bad github practice" but there were some changes that got forgotten in previous commits. whoops. it's all going to the same place anyways, as they say :)

bugs/sty: clean up comments & debug statements in init_mag_grid

5512844

FEAT: New input name, and default opts, for ionGrid.

e01ce55

This could break old input files, watch out! Seems ok on my end though...

BUG: Fix name of dAlt variable for ion grid in input files.

b0fca87

fix typo in grid.cpp

c1bcad2

STY: Continue enforcing local vs global vars having different names i…

c884617

…n mag_grid also move a couple functions around Just cleanin' up the code a bit

feat: calculate last latitude's p,q in fill_field_line

98f5831

minor linting & consistency things

f6d1222

"astyle, yo"

eeba18f

rm a func in grid.h from my debugging adventure

8596234

whoops. aether doesn't need this lol

Merge branch 'develop' into dipole_again

cd6447a

abukowski21 changed the title ~~Promised it would be done today, and it is.~~ Dipole grid for the ions Dec 20, 2024

abukowski21 marked this pull request as ready for review December 20, 2024 01:04

aaronjridley merged commit b429fe2 into AetherModel:develop Dec 20, 2024
3 checks passed

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Dipole grid for the ions #160

Dipole grid for the ions #160

abukowski21 commented Dec 18, 2024 •

edited

Loading

abukowski21 commented Dec 20, 2024

Dipole grid for the ions #160

Dipole grid for the ions #160

Conversation

abukowski21 commented Dec 18, 2024 • edited Loading

Dipole grid for the ions

Type of change

How Has This Been Tested?

Test configuration

Checklist:

abukowski21 commented Dec 20, 2024

abukowski21 commented Dec 18, 2024 •

edited

Loading