memo-keywords-across-time

Generate datasets with statistically over-represented words from a series of time periods, and visualize them in word clouds in a Shiny app.

Requirements

• Packages "tm" and "RWeka".

install.packages(c("tm", "RWeka"))

• A set of novels placed in the directory "corpustxts".

├── corpustxts
│   ├── 1870_AT_DenVanvittigesBoernEllerSkurkenPaaNoerrebro.txt
│   ├── 1870_AndersenHC_LykkePeer.txt
│   ├── 1870_Anonym_KraemmersvendeOgSyjomfruerEllerPrincipalernesPengeskuffer.txt
│   ...

• The directory "rdata".

Generate data

Data for the Shiny app are generated using "generate_keyness_data.R":

• Set working directory to the project directory.
• Make sure the directories "corpustxts" and "rdata" exist, the former filled with novel txts.
• Read all functions into R (i.e. run all code)
• Run make_data_chunks()

E.g.:

setwd("/Users/phb514/mygit/memo-keywords-across-time")
install.packages(c("tm", "RWeka"))
# Run all code (e.g. opt+cmd+R)
make_data_chunks()

Run Shiny app

• Copy rdata/signifdata.Rda to shinyapp/memo_keyness_over_time/rdata/signifdata.Rda
• Run library("shiny")
• Run runApp('shinyapp/memo_keyness_over_time') (or open app.R in RStudio and press "Run App").

Upload app to shinyapps.io

Google how to activate your free shinyapps.io account. Once you're set, run

rsconnect::deployApp('shinyapp/memo_keyness_over_time')

The app is now live at https://philip-diderichsen.shinyapps.io/memo_keyness_over_time