GGplot in Haskell

HaskellR.hs

demographicsAnalysis :: (MonadIO m, H.MonadR m) => Frame User -> m ()
demographicsAnalysis fr = do
  let demographics  = fmap (rcast :: User -> Record '[HaskellExperience
                                                     ,RecommendHaskell
                                                     ,Occupation])
                           
        (nubbed, counts) = view (runGetter ((,) <$> Getter (_1) <*> Getter _2))
                                               (unzip
                                                  (sortByLens (_1.recommendHaskell)
                                                              (sortByLens (_1.haskellExperience)
                                                                          (countAll demographics))))
      exp = L.fold (foldList (haskellExperience)) nubbed
      re = fmap fromIntegral $ L.fold (foldList (recommendHaskell)) nubbed :: [Int32]
      occ = fmap fromIntegral $ L.fold (foldList (occupation)) nubbed :: [Int32]
  liftIO $ do
    print $ length counts
    print $ length exp
    print $ length re
  [r| library (ggplot2)
      png (filename= "expVsrec.png")
      n <- length (exp_hs)
      print (counts_hs)
      print (length (counts_hs))
      print (n)
      dat <- data.frame(xvar = exp_hs
                       ,yvar = re_hs
                       )
      print (dat)
      print (ggplot (subset(dat, xvar>=-10)
                    , guide=FALSE
                    , aes (x=xvar, y=yvar)) +
               geom_jitter (aes (size=counts_hs, fill=factor(occ_hs), shape=21, alpha=0.2)) +
               scale_shape_identity () +
               scale_fill_discrete (name = "Occupation"
                                   ,labels = c("Developer"
                                              ,"Manager"
                                              ,"Student"
                                              ,"Faculty"
                                              ,"Other")) +
               # scale_size_area (max_size = 15) +
               scale_x_continuous (name="Haskell Experience", limits=c(-2,6), breaks=seq(-1,6)) +
               scale_y_continuous (name="Recommend Haskell", limits=c(-2,6), breaks=seq(-1,6)) +
               scale_size_area (name = "Population size", max_size=15)
               )
      dev.off() |]
  return ()