Progress Monitoring

[bbbccd]

There should be a way to monitor progress. For example

one option would be to add a parameter to print useful metrics to the screen

a parameter could be added to the initialization of the dk iterator, or to the synthesize method, which would cause progress to be printed on the screen

another option would be to add interface to allow the user to control progress through the iterations

For example:

for i in dk_iter.synthesize(...):
    print(i.peak_mu)

I tried using dk_iter.synthesize as an iterable as above, but I'm not sure how it works. In one place the docs say that an IterResult object is returned from synthesize, but the main example shows the results as a tuple K, N, mu, d_scale_fit_info, info. edit: ok i see the IterResult is inside the tuple. so in anycase you can't actually use it as in the above example to provide progress updates because it is not available until the synthesis is complete.

[sdahdah]

To some extent, this is already done via the logging library. If you change the log level like this

import logging
logging.basicConfig(level=logging.INFO)

in your script, you will see an output like

INFO:DkIterFixedOrder:Iteration: 0, mu: 1.1792755126953125
INFO:DkIterFixedOrder:Iteration: 1, mu: 1.0401535034179688
INFO:DkIterFixedOrder:Iteration: 2, mu: 1.0381889343261719
INFO:DkIterFixedOrder:Iteration: 3, mu: 1.0359954833984375
INFO:DkIterFixedOrder:Iteration complete

We will document this better.

IterResult is not an iterable (in the way you're thinking) so looping like that won't work. Maybe I could add a user callback that's called every iteration, but that's basically the same as this https://dkpy.readthedocs.io/en/stable/examples.html#d-k-iteration-with-a-custom-fit-order-selection-method

[bbbccd]

ok, thanks for the tip about logging.
maybe i was overreacting with this request, if most solutions are on the order of 5-10 iterations, and there is this logging workaround.
creating a custom DkIteration class is too much boilerplate.