Correct parameter for the subnet size in adaptive search

[snilsn]

This pull request aims at solving issue #133.

If neither adaptive_step nor adaptive_stop are given, MAX_SUB_NET_SIZE is set, otherwise MAX_SUB_NET_SIZE_ADAPTIVE is changed in the linking_trackpy()-function.

[freemansw1]

I've requested @JuliaKukulies and I as reviewers for this. I should be able to get to this next week.

Once this is reviewed, it may make sense to release it as part of a 1.3.2 bugfix release. I'm hesitant to try to get it into 1.3.1, but I could see this and #124 for example making it into a 1.3.2 release. For now, I've kept it on the 1.5 milestone, but may move it to a new 1.3.2 milestone.

[JuliaKukulies]

Thanks @snilsn, very good and careful observation! No objection from my side for the code changes you made. The only thing I wonder is whether we can add a more clear description for the subnetwork_size parameter in the docstrings (I mean to make clear that it means different things dependent on whether you specify adaptive_step/adaptive_stop or not). As I understand it, you would want it to be rather small when you do adaptive search and rather large when you do not. So this is important information the user needs to know.

[freemansw1]

I'm happy with these changes. Good catch on finding the bug.

[freemansw1]

With the approving reviews, @snilsn please confirm that you're ready to merge and we will get this in dev with the intent to release as part of v1.3.2.

[snilsn]

Yes, from my side everything is ready for merging.

The only thing I wonder is whether we can add a more clear description for the subnetwork_size parameter in the docstrings

I will adress this in an update of PR #138

[JuliaKukulies]

Good, yes agreed that the documentation update is better for #138 ! Thanks!