ialsa/ialsasynthesis

R Package for synthesizing information across collaborating research

ialsasynthesis's Issues

Visible Members

(Spun off of #5)

Looks like I'm the only one who's public (ie, visible to anyone who's not already a member), and I look lonely in this screenshot:

Consider asking the UVic (or anyone else) if they want to be visible. If not, just close the issue.

The settings for each person can be changed here.
https://github.com/orgs/IALSA/people

Coveralls

https://coveralls.io/ integrates with Travis-CI to determine how much of the functions are covered by tests.

• @andkov, can you please go to https://coveralls.io/ and turn on permissions for this repository? It's not giving me the options. I'm guessing it's because I'm not an admin of IALSA's GitHub "organization".
• after that, I'll follow the rest of the instructions in https://github.com/jimhester/covr

Add test output

Add a few output files from IALSA-2015-Portland to use as examples to test the parsing functions.

Descriptive README.md

(Assigned to @andkov and @wibeasley)
Add a description to the README.md file that will welcome users. Try to

• address many initial questions, and
• provide resources to other sites for the less common questions.

devtools::inst() is going away

This is head up to let you know that devtools::inst() will be removed in the next release of devtools. It looks you like you should be able to replace your use of it with system.file()

GitHub for package

@andkov, will you please

• See if you like the name IalsaSynthesis for this package?
• Assuming so, create a team called IalsaSynthesisPushPull
• Populate it with the people you want to have write privileges (of course this can be added to later).
• Tell me their public email addresses. This will be posted on CRAN.

prep for CRAN release 0.2.0

follows #28

Should we archive on CRAN

@andkov, any objection to archiving on CRAN? A dependency broke the builds on CRAN (#26). It's pretty easy to fix & resubmit, but I'm not sure there's much point. No one is using this package anymore, are they?

Note that the archived package can still be downloaded, as well as the live version installed from GitHub (with remotes::install_github("IALSA/IalsaSynthesis").

Create package structure

• DESCRIPTION file
• NEWS file
• refresh file
• Travis CI
• AppVeyor
• testthat

Zenodo DOI

• Add a DOI by Zenodo, so a page like this is independently archived: https://zenodo.org/record/13326#.VUE6yiFVhBc
• Add a badge to the README.md file that looks like:

images in README

excerpt of message sent by CRAN maintainer today:

...
Will Beasley [email protected]: IalsaSynthesis

These have README.md referencing local images which do not get shipped with the package, and hence missing in the README.html we generated from the README.md for the CRAN package web pages. See the list of missing image files below.

Can you pls fix to either ship the missing image files with the package (in which case you could put them in a suitable subdirectory of 'tools'
in case things like 'vignettes' or 'man/figures' are not appropriate), or use URLs (e.g. from your package source code repositories) instead?
...

$IalsaSynthesis
[1] "libs/images/ialsa_wide.png"

Submit to CRAN

Even though we'll need one more release before Gothenberg, I want to submit to CRAN this week for a few reasons, including

1. make the next submission smoother.
2. have the links working.
3. give SEO some extra time to mature.

Identify the elements to be extracted from .out model files

Blacklist

(Spun off from https://github.com/IALSA/Diabetes-Tri-Study/issues/1)

@andkov, do you have a plan for excluding them systematically & clearly? If there's a clear variable that explains why they should be dropped, do that programmatically (eg, a DOB before 1920)

If not, I suggest creating a "blacklist" called something like exclude_manual.csv or blacklist.csv. There's a row for every subject/observation/whatever that you want to drop with a manual specification. Typically I'll add a second column called something like exclude which is usually TRUE for most the rows, but it allows me to stop excluding them, while not deleting the record (and potnetially losing the ID values I was suspicious of). I'll also add a third column called notes that explains to outsiders why I didn't think they were a representative part of the sample.

Do you think many studies will have the need for this? If so, we might include some common functions in this package.

update package

to recent CRAN standards & package dependencies

• devtools::inst() is going away #26
• UTF-8 encoding
• testthat v3.0
• remove LazyData
  • https://developer.r-project.org/blosxom.cgi/R-devel/NEWS/2021/03/27#n2021-03-27
• remove documentation-peek I'm keeping it b/c there's no pkgdown site
• move from travis-ci.org to travis-ci.com
• move from NEWS to NEWS.md

Write/Transfer parsing functions

From IALSA-2015-Portland