Add a notebook for visualizing qc metrics

[JaerongA]

• Added quality_metrics.ipynb for visualizing qc metrics.
• Added documentation for each metric.
• Added pytest for ephys.QualityMetrics

[kabilar]

Thanks @JaerongA. Please make the following updates:

• Plots for all metrics in ephys.QualityMetrics.Cluster.
• Plots for all metrics in ephys.QualityMetrics.Waveform.
• Add x- and y-axis ticks and labels with units.

[kushalbakshi]

Hi @JaerongA, just a few small comments:

• At the start of the notebook, can you please add the following:

This notebook is used to visualize the spike sorting quality metrics using the metrics.csv generated from kilosort. 

If you are new to using DataJoint for array electrophysiology - please see our [tutorial](./tutorial.ipynb) notebook for an in-depth explanation for setting up and running the DataJoint Workflow for NeuroPixels electrophysiology.

The Element also offers Quality Metric visualizations. These are generated using metrics.csv generated from kilosort.

[kabilar]

Please update version and changelog.

[kabilar]

We are missing descriptions for the following metrics in the notebook:

• Nearest Neighbor Miss Rate
• Silhouette Score
• Max Drift
• Contamination Rate

[JaerongA]

We are missing descriptions for the following metrics in the notebook:

* [ ]  Nearest Neighbor Miss Rate

* [ ]  Silhouette Score

* [ ]  Max Drift

* [ ]  Contamination Rate

@kabilar I left some of these metrics blank because I couldn't find the information in the reference. Although some of them seem obvious, I didn't assume anything since the actual algorithm could be different. Can you provide me with a resource to fill in the gaps, if any?

[kabilar]

@kabilar I left some of these metrics blank because I couldn't find the information in the reference. Although some of them seem obvious, I didn't assume anything since the actual algorithm could be different. Can you provide me with a resource to fill in the gaps, if any?

Silhouette Score and Max Drift can be found here: https://elifesciences.org/articles/61834/figures#tables

[kabilar]

Thanks @JaerongA. To clarify, the last cluster metric plotted is Contamination Rate, not Cumulative Drift.

[kabilar]

Please add a description of Cumulative Drift.

[kabilar]

Could you please clarify the description for ISI violation.

ISI (Interspike interval) violation - Rate of ISI violation as a fraction of overall rate.

[kabilar]

@JaerongA I have submitted a few more suggestions to your fork/branch since it is currently difficult to provide suggestions on a Jupyter notebook: JaerongA#1

[kabilar]

Thanks @JaerongA! I have a few suggestions above.

[kabilar]

Thanks @JaerongA!

[kabilar]

Version 0.3.2 is tagged and released on GitHub.

[kabilar]

Docs have been released to render the quality metrics notebook on the website.