this project contains all the sources for the Capstone TextPrediction project for the Data Sciences Specialization on Coursera.

the direcory structure is as follows

• NGramLM.R : final S3 object for NGrams in R
• task 0..3: exploration dirctorsie with various R scripts
• task4
  • traintest.R: script to split corpora in ttaining and test sets
  • count.R: script to count Ngrams, counts stored in counts/
  • train.R: training scritpt that builds series of models with specifed parameters
  • evaluate.R measures coverage on evaluation corpus
  • evaluate2.R; measures perplexity on evaluation corpus
  • evaluate3.R: measures prediction accuracy on evaluation corpus, measurements saore in acc/
• quiz: scripts to provide answers to quizes
• python: python LM building (not used in final submission)
• MilestoneReport: Capstone 1.Rmd is the source for the report
• shinyapp: self contained directory with all resources for the Text predictor show case app
• FinalPresentation: FinalReport.RPres is the source for the presentation, others are data files
• FinalReport; NGramLM.RData contains source for final report (un-finished, not submitted)

not included

• data: all data downloaded for this project, and splt and subsampled versions
• models: directory with all models build