Goals for 9/27

[ebridge2]

@ebridge2

• [x] add covariance and mean prediction using old method (consider covariance everywhere a condition is present, and weight time points by the amount of signal present). Deliverable: same simulation results as last week, but this time with the new "initializer" method. NOTE: Not sure this will finish in time since I just added a LOT more goals to what to include in this round of simulation when talking with Jovo yesterday (see #43), so deliverable might be just code (took like 4 days to get results last time, so I don't anticipate it being done in time).
• [x] make markdown file with fMRI acquisition settings. Deliverable: https://github.com/NeuroDataDesign/fngs/blob/master/status/datasets.md.
• [x] find peeps to use my tutorials for functional connectome estimation (probably group members). Show that they can reproduce my mnr results starting with nothing but the repo and a link to download fMRI scan data from (ie, BNU1).
• [ ] review merging eric-dev-master with greg.
• [ ] send simulation results to kara and susan. Hopefully they will be excited. Again, see #43 with the reason why this isn't completed yet. We're getting really close but not quite there.

@ewalke31

• [x] implement standard EM in R
• [ ] if enough time, try to see where EM KF algorithm is going wrong
• [x] show in R markdown w/ examples

http://neurodatadesign.github.io/fngs/ewok_weekly/ewok_week_926.html

@ronnyb29

• [x] Do first chapter of probability primer (https://www.youtube.com/playlist?list=PL17567A1A3F5DB5E4) and take notes
• [ ] Help Eric with tutorials on functional connectome estimation and take notes if time permits

Not fully LaTexed due to MedHacks, link to scan of handwritten notes: https://github.com/NeuroDataDesign/fngs/blob/gh-pages/ron_weekly/sept26/PP_chapter1.pdf

@tagarwa2

• [x] Implement Kalman Filter in R http://imgur.com/a/h3bbi

[gkiar]

@ebridge2 @ewalke31 👌 ; @ronnyb29 add yours?

[ebridge2]

@02agarwalt

run on single subject of your choice. Provide me with plots, from R ggplot, of the downsampled timeseries file (y axis is the intensity, or the value, x axis is the time, and the grouping is by roi label), as well as a single correlation matrix from this subject.

http://imgur.com/a/1XZIw http://imgur.com/a/MDpcp

Show me mnr results for a single dataset that I provide. Let me know when you finish the above and I'll tell you how to get it.

@ewalke31

Run fngs pipeline on subset of BNU_2 dataset. The data is found at /mnt/ssd2/data_fmri/BNU_2. Run on all subjects found here.