DT_Bin2D.R

library(data.table)
data(movies, package='ggplot2')
library(lattice)

movies <- data.table(movies)

nBins <- 500

## 2D Binning
movies[, c('ratBin', 'votesBin'):=list(floor(rating/nBins), floor(votes/nBins))]
setkey(movies, ratBin, votesBin)
## Cross Join (as expand.grid) ...
bins <- CJ(0:nBins, 0:nBins)
## ... used to extract records from movies corresponding to each 2D bin
movies2 <- movies[bins, list(.N, rating, votes), nomatch=0]
## Finally aggregate with mean
movies3 <- movies2[,lapply(.SD, mean),
                   by=list(ratBin, votesBin)]

xyplot(votes ~ rating, data=movies3, cex=asinh(movies3$N)/5,
       alpha=0.6, pch=21, fill='blue', col='black')