Consistent approach needed for providing data to support code.

[tompollard]

Several of the chapters (for example, https://github.com/MIT-LCP/critical-data-book/tree/master/section2/chapter3c and https://github.com/MIT-LCP/critical-data-book/tree/master/section2/chapter3d) require associated data (in these cases, an arterial line dataset).

The data is sensitive and so cannot be included as part of the public repository. Instead we could provide:

• code for generating the dataset from MIMIC and/or
• sample data

[jraffa]

This is what I ended up doing for Chapter 5:

http://physionet.org/physiobank/database/mimic2-iaccd/

I don't know if that helps you or not. read.csv can source a url as an argument, so I load directly off of physiobank.

[tompollard]

That's a really nice approach...I'll try to set things up the same way.