Rotations are not defined for instrument 'nortek aquadopp'

[ddileonardo]

Describe the bug:

The dolfyn.rotate2 function does not work with aquadopp data. I get a ValueError: Rotations are not defined for instrument 'nortek aquadopp'. The api.rotate2 function shows the same error.

It seems like this data should be able to be rotated similarly to the Nortek Signature.

To Reproduce:

1. Read an aquadopp file
2. Set the declination with dolfyn.set_declination
3. Attempt to rotate the data
4. See the error

Minimal working example:
I suspect that this will show up with any aquadopp test file. This aquadopp was set to profile.

Code run was:

from mhkit import dolfyn
from mhkit.dolfyn.adp import api

declination = -1.86

data_path = "" #insert a profiling aquadopp file here

ds = dolfyn.read(data_path)

dolfyn.set_declination(ds, declination, inplace=True) 
api.rotate2(ds, "earth", inplace=True)

Expected behavior:

Expected the data to be rotated similarly to the signature instrument.

Desktop (please complete the following information):

• OS: Windows
• MHKiT Version: v1.1.2

[jmcvey3]

Ah, shoot, I should have made sure the tests would catch this. Because this is also a 1-line fix, @ddileonardo can you run through the rest of the analysis you'd like to do without rotating the data, and let me know if anything else fails as well?

[ddileonardo]

I can do that. I'll try to get to this by the end of the week.

[ddileonardo]

I've tried what I think is the final workflow for the aquadopp based on how I've processed some signature data from the same data collection.

The rotation is the only thing that has given me an error.

I was also unable to do the beam correlation plotting and cleaning based on beam correlation that I was able to do with the signature, but that seems to be because the aquadopp doesn't provide the beam correlation information. At least I don't see it in the dataset that results from reading the file.

[jmcvey3]

@ddileonardo Apologies, I should have caught this after fixing the first issue you brought up. A quick hack to bypass this problem is to set dataset.attrs['inst_model'] = "AWAC"; they use the same rotation schemes

[jmcvey3]

I've tried what I think is the final workflow for the aquadopp based on how I've processed some signature data from the same data collection.

The rotation is the only thing that has given me an error.

I was also unable to do the beam correlation plotting and cleaning based on beam correlation that I was able to do with the signature, but that seems to be because the aquadopp doesn't provide the beam correlation information. At least I don't see it in the dataset that results from reading the file.

Yes, the older Nortek instruments don't provide an acoustic beam correlation measurement. What you see after running mhkit.dolfyn.read() is everything that is stored in the raw datafile

[ddileonardo]

@ddileonardo Apologies, I should have caught this after fixing the first issue you brought up. A quick hack to bypass this problem is to set dataset.attrs['inst_model'] = "AWAC"; they use the same rotation schemes

Thanks! That is really helpful and will keep me moving forward.