3 beam solution

[jessecusack]

Sometimes one of my ADCP beams is hitting ice and needs to be excluded from the solution. This amounts to recalculating xyze.

To redo this we need to extract the following code from pycurrents for our own purpose:

            trans = Transform(angle=ppd.sysconfig.angle,
                              geometry=geom)
            if beam_index:
                ppd.vel[:] = ppd.vel[..., beam_index]
                ppd.amp[:] = ppd.amp[..., beam_index]
                ppd.cor[:] = ppd.cor[..., beam_index]
            xyze = trans.beam_to_xyz(ppd.vel, ibad=ibad)
            ppd.xyze = xyze

Line 1756 in rdiraw.py

It looks like bad beams can be specified! Unfortunately, there is no documentation on what ibad should look like...

I'll get to this after #17 is merged.

[gunnarvoet]

You can pass ibad (zero-based so beam 0 to 3) to Multiread(): https://github.com/gunnarvoet/gadcp/blob/6b49608b0935ef3b88d1b14b8a3c66cb57752999/gadcp/madcp.py#L301

[jessecusack]

That seems convenient. Do you know if ibad can be an array? i.e. the beam to exclude can vary with each ping?

The documentation implies it is just a single int.

[gunnarvoet]

Yeah, I don't think so.

[jessecusack]

Ok, that is less convenient. I think I'll have to do it by manually editing the vel array then. It should be as simple as replacing the bad beam data with its opposite beam.