For Nov. 28

[rmarren1]

Commit

Status	Task	Deliverable
🐼 :	New solution for Robust PCA	demo of the format of the code results in prep report @tsunwong123
:panda_face:	MARA algorithm (Complete Algorithm and Simulation described in Algorithm.md)	@mnantez1
:panda_face:	Fix the 'broken pipe' issue for webservice	an explaination the bug and how it was fixed @rmarren1
:panda_face:	dockerize webservice (and put on cortex)	next demo will be on cortex @rmarren1
:panda_face:	integrate kurtosis/power spectrum bad electrode detect	demo of results, working Kurtosis code.md notebook link and updated bad electrode detection, meanwhile the notebook located here to show it works @nkumarcc @rmarren1
:panda_face:	let user choose pipeline steps to use	demo going through it like user @rmarren1

Reach

Status	Task	Deliverable
:panda_face:	split git workflow into production and development	show branches
:x:	document dockerhub and functions	show docs @everyone
❌	start integrating prep pipeline and investigate other sources	beginning notebook @nkumarcc

[rmarren1]

@mnatenz1 @nkumarcc @tsunwong123 anything else?

[jovo]

@nkumarcc

• [ ] if your notebook is better than algorithm.md in certain ways, please tell me how so we can jointly improve
• [ ] do you know the formal definition of kurtosis/skewness?
• [ ] does your webpage have a description pipeline schematic illustration which steps you are performing?
• [ ] current pseudocode for "bad electrode detection"?
• [ ] what is the prep pipeline?

@tsunwong123

• [ ] what's the problem for @rmarren1 ?
• [ ] is there algorithms.md for your robust PCA?

@mnatenz1

• [ ] MARA: ugh. do they use it?
• [ ] is there any reason to suspect it is better (we can compare using algorithms.md with your existing method)

@rmarren1

• [ ] url to website?
• [ ] use neurodata.io webkit?
• [ ] use eeg.neurodata.io?
• [ ] include default for parameters

[ebridge2]

@mnatenz1 "Uses a machine learning algorithm..." which? Is it arbitrary? Should specify what is standard practice if thereare several.