Skip to content

Latest commit

 

History

History
111 lines (91 loc) · 3.53 KB

TODO.md

File metadata and controls

111 lines (91 loc) · 3.53 KB
Raw

Plan/To Do

Basic principles

  • shinydashboard layout
  • a module per tab to keep code easy to navigate/maintain
  • using fluid structure { fluidRow() -> column(width=..) -> shinydashboardPlus::box(width=NULL) } as the basis for every tabs ui
  • each box: single topic; valid title; solid header; consistent status/colours; collapsible & closable where desired
  • include code snippets where possible/appropriate, using the code_snippet() function (see R/fct_utils.R)
  • file naming conventions adhered to for any functions etc. that are required

Plan for each tab:

(Bullet points indicate rough/minimum points to cover, initials in brackets indicate who will complete)

  • Info (CB)
    • What the app is
    • How to use it
    • Important points

SECTION 1

  • What is Shiny, actually? (CB)

    • What a web app is
    • How shiny converts R to HTML, JS and CSS
    • Shiny dependencies (httpuv etc.)

  • Understanding reactivity (CB)

    • How reactivity makes a pull relationship look like a push relationship
    • Graph showing the flushing cycle and reactive graph
    • reactlog

  • Using reactivity (CB)

    • Types of object (reactive values, expressions, outputs)
    • When to use reactive vs observe
    • Including observeEvent vs eventReactive

  • Controlling reactivity (CB)

    • Using debounce & throttle
    • Using invalidateLater & reactivePoll
    • Using isolate & req

  • Structuring an app (CB)

    • As a package
    • File naming conventions
    • Workflow (using roxygen2, devtools, usethis etc.)
    • Making use of reactivevalues not being copy-on-modify
    • Using modules (briefly mention, covered in next tab)

  • [x ] Shiny modules (CB)

    • The 1000 button problem
    • Namespacing
    • Using modules effectively

SECTION 2

  • Improving UI with CSS

    • Key choices when designing shiny apps - choice between shinydashboard[Plus] or shiny+bslib (or other)
    • What is CSS and how to include in shiny
    • Using Sass

  • Improving UX with JS

    • What is JS
    • Ways to include JS in shiny
    • Some example use cases
    • Briefly mention shinyjs (covered in next tab)

  • The attaliverse

    • What is attaliverse
    • List all the packages
    • shinyjs
    • Some details of each, minimum include some detail for each of (shinyalert, shinycssloaders, shinydisconnect, shinybrowser)

  • Extending shiny (CB)

    • The power of javascript
    • Packages which extend/improve shiny (e.g. shinyhelper, shinyFeedback, shinyEffects)
    • Creating your own widget with htmlwidgets

  • Custom shiny inputs (CB)

    • Types of inputs in shiny & other packages
    • Input bindings and events
    • Creating your own input element
    • charpente

SECTION 3

  • The session object (CB)

    • Apps, sessions and users
    • Elements of the session object and what they are useful for
    • DfE-specific

  • Tracking usage

    • Why it can be useful
    • Custom methods (observers, save to sql etc.)
    • shinylogs

  • Advanced Visualisation 1 (plotly)

    • The basics of plotly
    • plotly vs ggplot2
    • Possibilities with plotly

  • Advanced Visualisation 2 (r2d3 + other js libs) (CB)

    • When to use r2d3 vs plotly/ggplot2
    • Extent of control
    • How to use js visualisations in shiny

  • Error handling & Debugging

    • base R error handling (tryCatch, withRestarts)
    • shiny-specific error handling (validate, safeError)
    • unit testing (briefly) and shiny testing (shinytest, testServer, testModule)
    • whereami

  • Optimisation

    • profvis
    • Caching
    • Async/futureverse