Open bsaul opened 2 months ago
This is related to #43.
I agree!
Another approach that we might consider (either in lieu of this proposal or in conjunction with) is to publish a website (using e.g. github pages) with the published csv files and perhaps some HTML file describing the available data and summaries/integrity checks of the data).
Another approach that we might consider (either in lieu of this proposal or in conjunction with)
I've been thinking about this approach for the past few days, and I think I like it. I might put up a PR that outlines the skeleton of what I'm thinking.
I might put up a PR that outlines the skeleton of what I'm thinking.
Tracking this effort in #90
Currently, the
mbbs
package is both a package containing functions for producing data file and a package of the data. This means, for one, the package that users might import for the data has a bunch of stuff and dependencies they don't need. For example, I shouldn't need to havebeepr
,readxl
,dplyr
, and so on installed just to load the mbbs data files.I propose that we keep the
mbbs
package as is, and then have a compilation step that produces another packagembbs-data
(?) that simply contains data files.