NeuroDataDesign / orange-panda-f16s17

Automated EEG data analysis.
5 stars 5 forks source link

For Nov 14th #21

Closed rmarren1 closed 7 years ago

rmarren1 commented 7 years ago

Commit

link to AWS instance

Status Task Deliverable
:x: Integrate Robust PCA into pipeline, make scalable demo of the results in prep report
:panda_face: Integrate ICA into the pipeline, make scalable, fix convergence issue, make parameters changeable via web service, and allow users to select and remove components demo of results in prep report
:panda_face: Implement and integrate coherence python notebook, script implementation and html accessible version
🐼 research kurtosis and power spectra in bad electrode detection latex and code.md and kurtosis in bad electrode
:panda_face: add token descriptions show page of token descriptions

Reach

Status Task Deliverable
:x: Automated independent component removal demo of the results in prep report
:x: split git workflow into production and development show branches
:x: dockerize webservice so we can scale up more easily dockerhub link
rmarren1 commented 7 years ago

@jovo we set a backlog (kanban style) here: https://github.com/NeuroDataDesign/orange-panda/projects/1

@mnatenz1 @nkumarcc @tsunwong123 want to do the poll here:http://doodle.com/poll/555sxkbw5s2xu49h#table to see when we can meet

jovo commented 7 years ago

1) responded to poll. email dude and see if any of those times work for him (of note, i can only do some of the times on wed). 2) backlog is cool, kanban style, but doesn't seem like you can link to issues in github projects?

On Mon, Nov 7, 2016 at 7:44 PM, Ryan Marren notifications@github.com wrote:

@jovo https://github.com/jovo we set a backlog (kanban style) here: https://github.com/NeuroDataDesign/orange-panda/projects/1

@mnatenz1 https://github.com/mnatenz1 @nkumarcc https://github.com/nkumarcc @tsunwong123 https://github.com/tsunwong123 want to do the poll here: http://doodle.com/poll/555sxkbw5s2xu49h#table to see when we can meet

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/NeuroDataDesign/orange-panda/issues/21#issuecomment-259010895, or mute the thread https://github.com/notifications/unsubscribe-auth/AACjcvJkpIgA-snvl-6LGiIixjxhHuwOks5q78YBgaJpZM4Kr5Dl .

the glass is all full: half water, half air. neurodata.io, jovo calendar https://calendar.google.com/calendar/embed?src=joshuav%40gmail.com&ctz=America/New_York

nkumarcc commented 7 years ago

@jovo: I think you can link to issues in github projects; if you look now at the bad electrode kurtosis and power spectrum task in the "In Progress" section, I was able to click it and create a Git issue that it linked to. I don't know how to link existing ones though

jovo commented 7 years ago

ok, everything in this issue should be a link.

rmarren1 commented 7 years ago

sent email to dude and CC'd you all. Also updated issue to link to other issues. I consolidated so might be missing details for your issues @nkumarcc @mnatenz1 @tsunwong123, maybe make sure everythings good?

jovo commented 7 years ago

this looks great! tbh, i would have written shorter task summaries for each, but definitely a massive improvement. any feedback you have on whether this seems better/easier for you i'd love to hear.

nkumarcc commented 7 years ago

@jovo: the only thing the other system had that was nicer was that we could more easily see who's responsibilities were what. Otherwise this system looks better for sure and in conjunction with the backlog makes more sense

rmarren1 commented 7 years ago

I like this organization. One thing I may try next week is linking directly to the 'cards' in the backlog page. That way we wouldn't need to create 8 different issues.

gkiar commented 7 years ago
ebridge2 commented 7 years ago

@nkumarcc as I said last night, would be dope if you make it a goal for this week to figure out the population detection algorithm being used to order the electrodes (ie, it looks like there is some sort of spatial organization of the elctrode numberings, but without knowing for certain that's just speculation; the higher coherences are all within a few nodes of the diagonal)

mlee156 commented 7 years ago

@nkumarcc I highly sympathize with things dying - if you host your jupyter notebook through a server you can have it publicly accessible anywhere!