NanoVNA connection syntax out of date in example Measuring a 4-Port With The 1.5-Port NanoVNA V2

[kpobrien]

Describe the bug
The example https://scikit-rf.readthedocs.io/en/latest/examples/metrology/NanoVNA_V2_4port-splitter.html#Measuring-a-4-Port-With-The-1.5-Port-NanoVNA-V2 is out of date and uses no longer functional syntax for operating the NanoVNA.

To Reproduce
Steps to reproduce the behavior:

>>> from skrf.vi import vna
>>> nanovna = skrf.vi.vna.NanoVNAv2('ASRL/dev/ttyACM0::INSTR')
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
AttributeError: module 'skrf.vi.vna' has no attribute 'NanoVNAv2'

Expected behavior
I expect to be able to successfully connect to the NanoVNA. Using the syntax from the comment at the end of #901, I'm able to connect successfully as in the example below:

from skrf.vi.vna import nanovna
instr = nanovna.NanoVNA('ASRL/dev/ttyACM0::INSTR')

System

• OS: Linux
• scikit-rf version: 0.32.0
• NanoVNA version: NanoVNA v2 Plus 4

Additional context
I'll submit a PR to update this example when I get a chance. Thanks!

[kpobrien]

The current (not working) version of the data acquisition portion of the example is:

import skrf
from skrf.vi import vna

# connect to NanoVNA on /dev/ttyACM0 (Linux)
nanovna = skrf.vi.vna.NanoVNAv2('ASRL/dev/ttyACM0::INSTR')

# for Windows users: ASRL1 for COM1
# nanovna = skrf.vi.vna.NanoVNAv2('ASRL1::INSTR')

# configure frequency sweep (for example 1 MHz to 4.4 GHz in 1 MHz steps)
f_start = 1e6
f_stop = 4.4e9
f_step = 1e6
num = int(1 + (f_stop - f_start) / f_step)
nanovna.set_frequency_sweep(f_start, f_stop, num)

# measure all 12 combinations of the 4-port
n_ports = 4
for i_src in range(n_ports):
    for i_sink in range(n_ports):
        if i_sink != i_src:
            input('Connect vna_p1 -> dut_p{}, vna_p2 -> dut_p{} and press ENTER:'.format(i_src + 1, i_sink + 1))
            nw_raw = nanovna.get_snp_network(ports=(0, 1)
            nw_raw.write_touchstone('./data_MiniCircuits_splitter/dut_raw_{}{}'.format(i_sink + 1, i_src + 1))

The latest acquisition code doesn't have the get_snp_network function, so I use get_s11_s21 then assemble the network manually. I tested the code below on a NanoVNA v2 plus 4 with scikit-rf 0.32.0 and it works.

import skrf
from skrf.vi.vna import nanovna
from skrf import Frequency

# connect to NanoVNA on /dev/ttyACM0 (Linux)
instr = nanovna.NanoVNA('ASRL/dev/ttyACM0::INSTR')

# for Windows users: ASRL1 for COM1
# instr = nanovna.NanoVNA('ASRL1::INSTR')

# configure frequency sweep (for example 1 MHz to 4.4 GHz in 1 MHz steps)
f_start = 1e6
f_stop = 4.4e9
freq = Frequency(start=f_start, stop=f_stop, npoints=1024,unit='hz')
instr.frequency = freq

# measure all 12 combinations of the 4-port
n_ports = 4
for i_src in range(n_ports):
    for i_sink in range(n_ports):
        if i_sink != i_src:
            input('Connect vna_p1 -> dut_p{}, vna_p2 -> dut_p{} and press ENTER:'.format(i_src + 1, i_sink + 1))
            s11, s21 = instr.get_s11_s21()
            nw_raw = skrf.network.four_oneports_2_twoport(s11, s21, s21, s11)
            nw_raw.write_touchstone('./data_MiniCircuits_splitter/dut_raw_{}{}'.format(i_sink + 1, i_src + 1))

A few questions:

1. Is this a reasonable way to generate the network?
2. An alternative is to implement get_snp_network in nanovna.py. I'm happy to do that once I understand (3).
3. Is my understanding that skrf/vi/vna/nanovna_v2.py is not functional (see the example in the bug report) and we should be using skrf/vi/vna/nanovna/nanovna.py for both v1 and v2? If that's correct, then I can port functionality such as get_snp_network from nanovna_v2.py to nanovna.py then delete nanovna_v2.py to avoid confusion.

Thanks!