R Miscellaneous Notes⚑

### Package Development Guidelines

The script should be used to document all changes and additions to package code. In addition, git-flow branches should be used for all added features and bug fixes. The **development** branch serves as the “test” / “work-in-progress” branch while the **master** branch is the “production” branch and should only be updated when creating a new **release**.

For example, say I wanted to add a new feature / function - the following steps would be used:

1. Using *Git-Flow*, a new branch should be created off the current **development** branch with a name corresponding to the new feature.
2. Once in the new branch, add functions via the **devhist.R** script via .
3. Edit the new .R file and commit changes and push to feature branch.
4. Add tests for new function and document additions via , and document this in the **devhist.R** script.
5. Add example usage of the new feature via , and document this in the **devhist.R** script.
6. Document, Test, Check, Build, and Install.

7. Finally, merge feature branch with development branch using git-flow and if desired create a new release and merge with master branch updating the package version.

   Ideas⚑

   Functions⚑

Workflow Functions⚑

• Project Configuration
• Version Control and Git
• Directory Structure
• Project Variables
• External Data Sources
• Caching
• Pipelines
• Databases and Query’s
• GoogleDrive
• Files Metadata
• Codebooks, Data Dictionaries, Name Tables
• Documentation
• CI
• Testing
• Code Review

Actuarial Functions⚑

• Compare to Prior
• Retentions
• Policy Periods
• Grouping by Occurrence
• Exposure Lagging
• Experience Mod Factors
• Industry Loss Costs
• Data Validation Checks
• Manual Adjustments
• Triangle Formation
• LDF Derivation
• LDF Interpolation
• Discount Factors
• Data Summaries
• Loss Costs
• Severity
• Frequency
• Credibiilty
• Allocation
• Projections
• Development Methods
• BF Methods
• Parrallelogram Methods
• Evaluation Dates / Year-Month Time Series Data
• Persistency
• NCCI
• Credibility
• Simulations and Confidence Intervals / Levels
• Benchmarking
• Contigency Tables
• Survival Analysis

Utility Functions⚑

• Parsing
• Dates
• Excel

Data Structures⚑

• Lossruns
• Triangles
• Exposures
• Industry Loss Costs

RStudio Addins⚑

• Project Init
• Dependencies
• Styling
• Create Script with Header
• Git Setup
• Name Table Addin
• Excel Integration Addin(s)
• Remote / Raw Data Directories (Persistent Data Storage)
• Shiny App Init
• Find and Replace
• Data Manipulation (MiniUI with a Pivot Table)
• Markdown Addins
• Prefixer / @importFrom
• Clipboard

Templates⚑

• Project Directory
• Package Directory
• Shiny App Directory
• Markdown HTML Template
• README Template
• Package DESCRIPTION template
• Excel Output Template
• Shiny Apps HTML Templates
• HTML Widget Templates
• PDF Acturial Report Template
• Latex Tables

Shiny Apps⚑

• Modules
• Highcharter
• DT
• CSS, Bootstrap, Java, HTML

Options⚑

• Global options
• .Rprofile
• .Renviron
• Git
• BB & GH
• Conflicts
• DT
• Knitr
• Pre-Commit Hooks
• Shiny
• Highcharter

Data⚑

• CRM - People, Emails, Phones, Expertise, Clients, Budgets, etc.
• Historical Archived Data (Versioning)
• Git LFS
• Example Lossruns, Exposures, Industry Data
• Triangles, LDFs, Discount Factors
• Simulation Outputs
• Transactional Data
• Liberty, Sedgwick, etc extraction.

Resources and Learning⚑

• https://rtask.thinkr.fr/when-development-starts-with-documentation/
• https://emilyriederer.netlify.com/post/rmarkdown-driven-development/
• https://r-pkgs.org/index.html
• https://github.com/ThinkR-open/golem
• https://thinkr-open.github.io/building-shiny-apps-workflow/
• https://github.com/ThinkR-open/chameleon
• https://github.com/rorynolan/exampletestr
• https://github.com/ThinkR-open/attachment/blob/master/devstuff_history.R

---

Links: Source: