Inherited from ukgovdatascience/rap_companion#80

Relevant: https://www.gov.uk/guidance/content-design/writing-for-gov-uk

https://ukgovdatascience.github.io/rap-website/article-dependency-and-reproducibility.html

In progress by @sebastian-fox -- thanks!

• Collaborative document editing with redoc or maybe markdrive.
• Publishing analysis as HTML vignettes
• Publishing analysis as HTML govdown
• officedown
• make spreadsheets with writexl openxlsx

• [email protected]
• meets monthly
• Sort the C&C orgs alphabetically

Thanks @StuartNewcombe for reporting

Add a page to resources that links to style guides

• https://datasciencecampus.github.io/coding-standards/
• https://moj-analytical-services.github.io/our-coding-standards/web/
• https://www.isdscotland.org/About-ISD/Methodologies/

Please add more links in the comments.

Currently Travis deloys to the gh-pages branch using a personal account. We should create a new account to be used for this kind of thing.

https://developer.github.com/v3/guides/managing-deploy-keys/#machine-users

Also needs documentation here and maybe in https://github.com/ukgovdatascience/govdown.

Thanks @tjtnew for reporting.

I propose we tweet a lot about RAP using the handle @RAP4Stats

Add DfT Coffee & Coding email: [email protected]

Build on push to master?

Examples:

• https://github.com/adamrobinson361/r-workflow/blob/master/.travis.yml
• https://github.com/departmentfortransport/ds-r-training

Also write a how-to article.

How can we make it easy for:

• RAPpers to work towards the best possible form of RAP for their publication?
• RAPpers to seek help from others who are further down the RAP path?
• users to have an at-a-glance understanding of the level of reproducibility and automation used to create a publication?

This could be achieved by (self-)assigning relative 'levels'. Producers at 'level 2' could seek help from those at 'level 3' and above, for example. I think it's more important for collaboration and help-seeking internal to government, rather than creating a possible badging system (e.g. 'RAP level 5').

The companion seems like a sensible place for this to exist, but I would expect that it would really be in the hands of GSS to investigate and organise anything formal.

Anna Price, NHS National Services Scotland, produced a publication about RAP that contains a table describing seven levels of code maturity and automation that could be generalised: https://www.isdscotland.org/About-ISD/Methodologies/_docs/Reproducible_Analytical_Pipelines_paper_v1.4.pdf

Include links to publications/analysis/outputs which were produced using RAP including a link to the code if publicly available.

Since RAP concepts should be language agnostic I propose that we need to add more Python content.
I know that there are quite a few analysts in the government that are using Python for their RAP work so this is going to be widely useful.

Discussion thread for the infrastructure page. See also the original discussion. Please comment here to suggest:

• adding new tools
• removing obsolete tools
• rethinking some advice
• etc.

As the guidance is live and available to external users, would it make sense to change edit the ALPHA banner to BETA instead? @nacnudus

For discussing the Coffee & Coding page. Please comment here to

• add a group
• add details about a group
• improve the blurb
• etc.

Treat it as your wiki.

I'm seeking volunteers to write about Coffee & Coding for the Reproducible Analytical Pipelines website.

Three levels of contribution:

1. Easy: Add C&C clubs you know of to this issue.
2. Easy for you, hard for me: Write a paragraph or so that explains what C&C is and encourages people to participate.
3. Learning opportunity to try markdown and GitHub: make a pull request with a markdown version of the page.

Thanks!

Original issue description: "Packrat is more trouble than it's worth: discuss."

The packrat package is for managing dependencies. Informal comments at the meetup suggested it can be a pain, especially on Windows when attempting to compile old versions of packages. Docker was suggested as an alternative with the benefit that it isn't specific to R.

Discussion (packrat-focussed): https://github.com/ukgovdatascience/rap_companion/issues/86
More Discussion: ukgovdatascience/rap_companion#89
Article in the RAP companion: https://ukgovdatascience.github.io/rap_companion/dep.html (PR)

@matt-dray I think your RAP Companion chapter would be a good standalone article. It's on the RAP website already as part of the companion, but I think it should be moved. What do you think?

Link to BEIS QA methodology, Aqua Book, GSS Good Practice Team, "Inter-departmental working group on quality assurance of analytical models", rOpenSci.

Something about whether to re-implement models as a check.

Anything else people suggest on Slack.

The main motivation behind this article is for helping people, especially non-technical managers, why you might want to take a more sophisticated approach to a RAP project, than say simply writing a single R script.
See PR #8.

https://www.udemy.com/reproducible-analytical-pipelines/

Guidance on "how to evaluate RAP" and build evaluation into the whole process

All new public sector websites will need to meet accessibility standards and publish an accessibility statement.

Accessibility in Government blog post: https://gds.blog.gov.uk/2018/11/21/public-sector-website-accessibility-statements-what-you-need-to-know/

I.e. precompiled for when you can't compile locally.
https://ukgovdatascience.github.io/rap-website/article-dependency-and-reproducibility.html

Open to ideas....

What are the minimum/nice-to-have tech specs for RAP (hardware/software)?

ROpenSci: https://ropensci.github.io/dev_guide/