How to plot and install?

[eabase]

@westphae
How can we use this to make plots?
Alternatives?

@zbednarke
I see you made an updated fork, can you please enable the issues for your repo?

[westphae]

Two questions in one — splitting them:

Install

Now covered in the README:

go install github.com/westphae/geomag/cmd/wmm_point@latest    # CLI
go get github.com/westphae/geomag@latest                      # library

Requires Go 1.22+. WMM2025 coefficients are embedded; no separate data file.

Plotting

geomag doesn't include a plotting layer — it only computes field values. There are a few ways to get plots, ordered roughly by effort:

1. Use NOAA's pre-rendered maps. NOAA publishes maps with each WMM release: https://www.ncei.noaa.gov/products/world-magnetic-model → "Maps". You'll find PDF/PNG plots of declination, inclination, total intensity, etc. for the current model. If you just need standard contour maps these are the canonical source and zero work.

2. Plot in Go with gonum/plot. For programmatic plotting in Go: compute the field on a lat/lng grid and feed into gonum.org/v1/plot. Sketch:

import (
    "gonum.org/v1/plot"
    "gonum.org/v1/plot/plotter"
    "github.com/westphae/geomag/pkg/egm96"
    "github.com/westphae/geomag/pkg/wmm"
)

m := wmm.Default()
t := time.Now()
// Build a grid of declination values
data := make(plotter.XYZs, 0)
for lat := -80.0; lat <= 80; lat += 5 {
    for lng := -180.0; lng <= 180; lng += 5 {
        loc := egm96.NewLocationGeodetic(lat, lng, 0)
        f, _ := m.MagneticField(loc, t)
        data = append(data, plotter.XYZ{X: lng, Y: lat, Z: f.D()})
    }
}
// then draw a heatmap or contour from `data` per gonum/plot's docs.

gonum/plot does heatmaps natively and contours via the plotter.Contour type. Adding a coastline overlay is harder — gonum doesn't ship one.

3. Plot in Python. For map-style plots with coastlines, projection support, etc., the path of least resistance is to dump computed values to CSV from a small Go program and feed into Python's matplotlib + cartopy. This is the same workflow most published WMM-related papers use.

4. Web-based (D3/leaflet). If you want interactive in a browser: emit JSON from the Go side, render with D3 or a Leaflet heatmap plugin.

I'd start with option 1 unless you specifically need a plot keyed to your own grid/coordinates/date — in which case option 3 is usually the fastest path to a publication-quality figure.

Re: enabling issues on @zbednarke's fork — that's outside this repo, you'd need to ping them directly or open an issue against their fork once they enable the tab.

Closing as the install half is now documented and the plotting half is answered above. Reopen if you want to dig deeper into any specific plotting path.

[eabase]

@westphae

Great! I need to come back to this topic, as it's been a while...

• As for your AI suggestion of using #1 that is a no go. I can't use NOAA (or NASA) pre-doctored material, and need reproducible sources and plots.

• #3 (and perhaps #4) sounds most promising!
  Can you ask your AI be a little more specific? Perhaps help with some demo scripts?

I Can't reopen issue, as you have not enabled that permission.

[westphae]

@eabase reopened for you. If you can be more specific about what you would like to plot, I can perhaps help you out a little (with AI help ;P ) using one or the other approach. Some Python script would be super easy.