Skip to content

Latest commit

 

History

History
104 lines (84 loc) · 4.04 KB

future_developments.md

File metadata and controls

104 lines (84 loc) · 4.04 KB
Raw

Future Tutorial Ideas

Fortran - Declan

More Python - Declan and James

Webscraping using Python - Izzy

Structural equation modelling - Pedro / John?

Data vis 3 - animated graphs - Gergana

Visualising movement, maps with lines connecting dots - Gergana

Interactive maps using leaflet - Gergana

Cleaning up datasets in more detail - Gergana

  • Subsetting using []
  • Renaming rows and columns
  • Deadling with spaces in rows (e.g. when you think it's a NA, but it's not blank, cause there is a space)
  • Getting rid of NAs
  • Reordering levels of a categorical variable
  • Re-ordering graphs based on e.g. ascending, descending order of a variable

Using the vegan package for similarity & composition analysis - Gergana

  • Turning a species list into a presence/absence matrix
  • Calculating Simpson's, Shannon, Jaccard, Sorensen and Chao indices
  • PCA & ordination analyses

Some people have suggested a workshop specifically on base R tricks

  • Subsetting using [] - Note this is currently in "Cleaning up datasets in more detail - Gergana"
  • for loops, if/ifelse, functions - In John's functional programming intro
  • lists

Build on troubleshooting tutorial

  • Building a Minimal Working Example to ask a question on Stackoverflow
  • Reading R help files
  • Link to the /links page on the website

Power analysis - Gergana

Reproducible research - Gergana

  • Building on Markdown and etiquette tutorial
  • Formatting graphs and tables for professional publication
  • packrat - for package management in a project
  • roxygen2 for making a personalised R package
  • Storing your data in a package

Dealing with Large Datasets - John

  • Import methods
    • data.table
    • fread
    • read_
  • Data Manipulation
    • Being parsimonious
    • Taking advantage of R object orientated environment
    • Pipes revisit
  • .RData objects
  • Saving a workspace

General RStudio IDE Customisation

  • Addins
  • Syntax highlighting
  • Autocomplete

Scenario challenge

  • Test student's abilities by drawing on ideas from multiple tutorials
  • In the context of taking an experiment through data exploration, analysis, presentation, publishing
  • Example:
    • Clean and get ready for analysis (Tut.2)
    • Create explanatory figures of relationships between variables (Tut.4)
    • Test a hypothesis using general linear models (Tut.5)
    • Test another hypothesis using a heirarchical model (Tut.7)
    • Create an Rmarkdown script of their work (Tut.6)
    • Upload the whole lot to a github repo (Tut.3)

For Sept and onwards, Gergana will draft an email asking people if they want to write a tutorial for the website and run a CC workshop in their institution, we can then all send that to the academics we know. And also check who is coming for seminars, maybe they have stats knowledge to share.

Website developments

  • Expandable dropdown boxes for parts of tutorials, to streamline tutorial page
  • Common problems page (could this be an interactive flowchart?)

Future CC developments

  • Build links with Python people, hold a Python session - [email protected]
    • Boris Gailleton (Geomorphology Masters student starting PhD Sept 2017)
  • Find someone to teach a Java session - [email protected]
  • Collaborate more with Drummond people - Geo taster sessions - [email protected]

Promotion

  • In Sem 1, put up posters around KB, Central and Drummond - Gergana
  • International posters - Gergana, done!
  • Speak at the info sessions for 1st year EES, the info session for 4th year EES
  • Email lecturers to ask them to tell students about Coding Club, esp in their meetings with tutees
  • Email EVERYONE to tell them about CC and ask them to join our mailing list if they want

Future years

  • Expanding Coding Club to Linguistics and Psychology?
    • Books
      • Lewandowska - Linguistics
      • Andy Field - Psychology
    • Market
      • Hons diss students - Feb/Mar
      • MSc diss students - May-July
      • Both get a short intro but only a minority are quantitative minded so nothing indepth given
      • They can consult Tom Booth (Psy lecturer) but normally this lecturer isn't available when they really need it