11. Integration tests against `hyperdrive simulate`

[JLBLine]

Final tests can be used to check the correct order of dipole flags and amplitudes and being. These won't be part of an automated test as it relies on running hyperdrive, but used to check everything that's been coded up. Change anything from previous parts of milestone to ensure these tests work. Do the following:

• Simulate a single point source, not at phase centre, without the primary beam, and compare visibilities. This way we can ensure output visibilities are in the same order, and that we're comparing baseline to baseline in correct order.
• Turn on the FEE beam, check tolerance for how similar the visibilities come out without dipole flagging and amplitudes
• Switch on dipole flags and check things match
• Switch on dipole amplitude and check things match
• Switch on both flags and amplitudes and check things match

[JLBLine]

Things to look out for when comparing with hyperdrive:

• when defining a visibility, for u,v,w, is it antenna 1 - antenna 2 or antenna 2 - antenna 1. This is important because the primary beam value is conjugated for antenna (tile) 2. So we could be grabbing exactly same beam values but applying them differently. Check the d_ant1_to_baseline_map, d_ant2_to_baseline_map arrays

[JLBLine]

Added an example in WODEN/examples/dipole_ampflags in 9ba45363b02add3841581df12dfba66cea7146ea that directly compares WODEN to hyperdrive simulate, checking the ordering of dipole amplitudes and flagging match.