[Feature] Online batched statistics measures

[dizcza]

Welford's online algorithm for the population mean and variance estimation. Partially addresses #151. The primary idea of this feature, however, is to use the batched computation approach as a part of Elephant-NEST and Elephant-NEST Desktop integration.

The pull request can be merged at any moment - the basics have been implemented.

• [x] connected to project of @ojoenlanuca tbd

[coveralls]

coverage: 88.4% (+0.1%) from 88.264% when pulling bf80d4867673ece96d3af46b6d8a4304b5f7ce7a on INM-6:feature/online_measures into 9326804ac14636d4d192c26bfdf1bf8d4bc88977 on NeuralEnsemble:master.

[Moritz-Alexander-Kern]

This PR seems to be related to @ojoenlanuca s project.

Link: https://github.com/ojoenlanuca/online_elephant

[pep8speaks]

Hello @dizcza! Thanks for updating this PR. We checked the lines you've touched for PEP 8 issues, and found:

There are currently no PEP 8 issues detected in this Pull Request. Cheers! :beers:

Comment last updated at 2024-04-04 13:07:13 UTC

[Moritz-Alexander-Kern]

On a general note, how about creating a subpackage for all online measures ? e.g. :

elephant
├── online
│   ├── __init__.py
│   ├──  mean_online.py
│   ├──  variance_online.py
│   ├──  inter_spikeInterval_online.py
│   ├──  ...
│   ├──  unitary_event_analysis_online.py

With PR #491 online.py will grow and with more online measures to come in the future, this could be a sensible way to structure this.

Update: 07.11.2022

• we agreed to not implement this, as it is not well suited for the class based structure of online analysis.