corr.R This file is used for the John Hopkins Data Science Specialization (R Programming). This file is posted for the blog post reviewing the specialization https://medium.com/@GalarnykMichael/in-progress-review-course-2-r-programming-jhu-coursera-ad27086d8438#.38n89lga5

corr.R

corr <- function(directory, threshold = 0) {
  
  # Reading in all files and making a large data.table
  lst <- lapply(file.path(directory, list.files(path = directory, pattern="*.csv")), data.table::fread)
  dt <- rbindlist(lst)
  
  # Only keep completely observed cases
  dt <- dt[complete.cases(dt),]
  
  # Apply threshold
  dt <- dt[, .(nobs = .N, corr = cor(x = sulfate, y = nitrate)), by = ID][nobs > threshold]
  return(dt[, corr])
  }

# Example Usage
corr(directory = '~/Desktop/specdata', threshold = 150)