Version 2.2 Release

[ryanhammonds]

I've aggregate/organized the changes since the last release to help with the 2.2 changelog and to track what's left to be done before the next release.

@TomDonoghue feel free to change this as you see fit. I think wrapping up open PRs marks a good target to release 2.2, but feel free to add additional to-dos here.

Aperiodic

• Detrended Fluctuation Analysis ([ENH] - Add DFA to a new aperiodic module #167)
• IRASA ([ENH] - Add IRASA #212)
• Autocorrelation ([ENH] - Add autocorrelation measure #217)

Simulation

• Burst detection bug fix ([BUG] Fix detection of burst on- and offsets. #220)
• Timeseries with defined spectral parameters (i.e. height, width, center) ([ENH] - Simulation of Central Frequency in the Time Domain #221)
• Timeseries with defined Lorentzian parameters (i.e. chi1, chi2, knee) ([ENH] - Simulate time series with knees, sum of sinusoids #222)
• Define bursts based on set durations ([ENH] - Update approach for simulating bursts #239)
• Extrema phase shifting (i.e. trough-to-trough or peak-to-peak) ([ENH] Min-to-min and max-to-max phase shifting #247)
• Custom cycle support ([ENH] - Add support for passing in a callable to sim_cycle #248)
• Cycle length fix ([BUG] - Update standard for compute n_samples for simulations #250)
• Variable oscillations parameters (i.e. vary cycle simulation parameters on a cycle-by-cycle basis) ([ENH] Simulate variable oscillations #252)
• One-sided asymmetry ([ENH] - Add one sided asymmetric option to asine cycle #254)
• Simulate fractional Gaussian noise and fractional Brownian motion ([ENH] - Simulate fractional Gaussian noise and fractional Brownian motion. #216)
• Asymmetrical gaussian cycles ([ENH] - Asymmetrical gaussian cycles #228)
• Simulate spikes of action potentials ([MRG] Simulate spikes / action potentials #259)

Plotting

• 2d array plotting (timeseries: [ENH] 2d Array Plotting #246, spectra: [ENH] Plot spectra accepts 2d arrays #269)
• Labeling fix ([MNT] - Update tests & small plot bug fixes #255)
• Improved tests ([MNT] - Update tests & small plot bug fixes #255)
• Plotting saving updates ([ENH] - Update default for savefig #258, [ENH] - Extend access to options for saving figures #260)
• Plot bug fixes ([MNT] - Update tests & small plot bug fixes #255)

Maintenance

• Deprecate increase_n argument of robust_hilbert ([FIX] Deprecated hilbert_increase_n arg #215)
• Error for invalid transition bands ([FIX] Error for invalid transition bandwidth #267)
• Improved speed of multidim decorator ([MRG] Add multidim decorator to filter_signal #264)

Documentation

• Tutorial for Morlet wavelets ([DOC] - Tutorial for morlet wavelets #226)
• Tutorials for DFA, IRASA, and Autocorrelation ([DOC] - Add tutorials for aperiodic sub-module #231)
• Tutorials and updates for lagged coherence and sliding window matching ([ENH] - Update rhythm code & docs #230)

[TomDonoghue]

This looks great, thanks for doing this @ryanhammonds!

I don't we want to add anything beyond the current open PRs for 2.2, as soon as we get the open things merged we can do a new release.

[ryanhammonds]

Hi @TomDonoghue, I caught up today on all the open PRs and they should be ready to go. I'll wait to merge the open PRs until you get a chance to approve them too. We've both already reviewed them so hopefully it won't be too much work. After that, I'll start preparing a new release!

[ryanhammonds]

Hi @TomDonoghue, what do you think about removing #235 from 2.2 and getting a release out soon? I was going to release a new bycycle version but noticed it requires neurodsp 2.2 and it seems better to release with a neurodsp pypi dependency rather than a link to the main branch.

[TomDonoghue]

@ryanhammonds - yes, I think that is a good idea! We should just get a new release out at this point, and there's no reason not to move #235 into a 2.3 release! Sorry for dropping off on working on the remaining PRs to get the release out!

The last open questions then seems to be whether there are things in #275, #277, and #303 that should be prioritized into 2.2. I think #277 can merge in quick, but the other two seem to have some open questions - can you do an a quick status check and see what you think should happen with them? If you tag me in, I'll do my best to help with anything over the next few days.

[ryanhammonds]

Yeah, I think #277 can be merged as is. #275 and #303 are smaller bugs / edge cases that can wait until 2.3 or maybe a smaller minor update (2.2.1).

So I can merge #277 (lmk if you want to check over it first) and then go ahead and release if that sounds good to you?

[TomDonoghue]

This sounds like a good plan to me! I agree that #275 and #303 are minor enough points to be merged into a future release.

I had a quick check of #277, and that looks good to merge!

There is also #299 open, which most notably adds the 'aperiodic' module to the list of sub-modules. When you are ready to release, can you have a quick check for anything else worth adding into the readme, and then merge that?

Also, also - it crossed my mind that testing against 3.10 should now be supported from Github actions, so I have added that in #307. As long as that looks good to you, do you want to merge that in as well, and then 2.2 officially supports 3.10?

Otherwise I think we are ready for a new release - are you good to go with that after merging these last things in? Other than using this issue to list the updates on release page, I think it should be a normal release! Let me know if you need anything else from me - but there's nothing else I want to check in on, so also feel free to just go for it!