Using R Markdown to automate Historical Census management

Using R Markdown to automate Historical Census management information Harrison Davies – Office for National Statistics

What is R Markdown? ► Markdown – Plain text syntax ► Shiny – Interactive graphics ► Dplyr – Manipulation ► Knitr – Output ► Ggplot 2 - Graphics

How is it useful? ► Markdown – Flexible format ► Shiny – Useful interactivity ► Dplyr – Intuitive data manipulation ► Knitr – Range of different outputs ► Ggplot 2 - Intuitive graphic creation

My experience with R Markdown MData. Gov modules using R Markdown: ► Data Science Foundations ► Statistical Programming ► Regression Modelling ► Introduction to Machine Learning

Overview - Team Customer Services Data Management Historical Census team Research Commissioned Tables

Overview - MI Customer Services Data Management Historical Census team Research Commissioned Tables

MI process Commissioned Tables Customer Services Research

How can R Markdown help? It can help: ► Unify the report by pulling in multiple data sources ► Automate ► Make the manual excel steps graphics more appealing ► Provide more detail with interactivity ► Enhance the context through graphics

CS - Task Charts

CS - Feedback Charts

CS - Feedback Comments

Research – Task Charts

Research – Recently completed and Pipeline

Useful code ► rmarkdown: : render("MANAGEMENT_INFORMATION_MARKDOWN. Rmd") ► plotly: : ggplotly(CS_FB 5) ► select. list(format(Sys. Date() - months(0: 12), "%Y-%m"), graphics = TRUE) ► DF %>% tidyr: : separate(Topic, LETTERS[1: 22] , sep = ", ")

Any questions?