column merge and split with selecting columns

library(pipeR)
library(plyr)
library(data.table)
library(dtplyr)
library(dplyr)
library(tidyr)

# data generation
num_csv <- 3e3
num_xlvls <- 30
num_ylvls <- 30
timePoints <- expand.grid(paste0("a", 1:num_xlvls),
                          paste0("b", 1:num_ylvls),
                          stringsAsFactors = FALSE)
csv_files <- lapply(1:num_csv, function(i){
  dat <- sample(1:nrow(timePoints), sample(400:600, 1)) %>>%
    rep(each = 5) %>>% (timePoints[., ]) %>>%
    modifyList(setNames(lapply(1:3, function(j){
      sample(1:100, nrow(.), TRUE)
    }), paste0("Var", 3:5))) %>>% tbl_df
})

selectedCol <- c(3, 5)

st <- proc.time()
outRes2 <- lapply(csv_files, function(subdf, pasteCols){
  aggFml <- as.formula(paste0("~paste(", paste0(pasteCols, collapse = ","), ",sep=';')"))
  tmpdf <- subdf %>>% mutate_(.dots = setNames(
    list(lazyeval::interp(aggFml, .values = environment())), "tmp")) %>>%
    group_by(Var1, Var2) %>>%
    summarise(values = paste(tmp, collapse=';')) %>>% ungroup
  numCols <- length(strsplit(tmpdf$values[1], ";")[[1]])
  tmpdf %>>% separate(values, paste0("V", 1:numCols)) %>>%
    mutate_each(funs(as.integer(.)), -Var1, -Var2)
}, pasteCols = paste0("Var", selectedCol)) %>>% bind_rows
proc.time() - st
#  user  system elapsed
# 66.91    0.09   67.75


st <- proc.time()
outRes3 <- lapply(csv_files, function(subdf, pasteCols){
  evalStr <- paste0("paste(", paste0(pasteCols, collapse = ","), ",sep=';')")
  tmpDT <- subdf %>>% data.table %>>% `[`( , tmp1 := eval(parse(text = evalStr))) %>>%
    `[`( , list(tmp = paste(tmp1, collapse = ";")), by = c("Var1", "Var2"))
  numCols <- length(strsplit(tmpDT$tmp[1], ";")[[1]])
  tmpDT %>>% `[`( , `:=`(paste0("V", 1:numCols), tstrsplit(tmp, ";"))) %>>%
    `[`( , tmp := NULL) %>>%
    `[`( , `:=`(paste0("V", 1:numCols), lapply(.SD, as.integer)),
         .SDcols = paste0("V", 1:numCols))
}, pasteCols = paste0("Var", selectedCol)) %>>% rbindlist
proc.time() - st
#  user  system elapsed
# 36.70    0.89   37.32

st <- proc.time()
outRes4 <- lapply(csv_files, function(subdf, pasteCols){
  aggFml <- as.formula(paste0("~paste(", paste0(pasteCols, collapse = ","), ",sep=';')"))
  tmpdf <- subdf %>>% tbl_dt %>>% mutate_(.dots = setNames(
    list(lazyeval::interp(aggFml, .values = environment())), "tmp")) %>>%
    group_by(Var1, Var2) %>>% summarise(values = paste(tmp, collapse=';')) %>>% ungroup
  numCols <- length(strsplit(tmpdf$values[1], ";")[[1]])
  tmpdf %>>% separate(values, paste0("V", 1:numCols)) %>>%
    mutate_each(funs(as.integer(.)), -Var1, -Var2)
}, pasteCols = paste0("Var", selectedCol)) %>>% bind_rows
proc.time() - st
#  user  system elapsed
# 55.94    1.34   57.13

# print out the versions of every used pkgs
# sessionInfo()
# R version 3.2.5 (2016-04-14)
# Platform: x86_64-w64-mingw32/x64 (64-bit)
# Running under: Windows 7 x64 (build 7601) Service Pack 1
#
# locale:
# [1] LC_COLLATE=Chinese (Traditional)_Taiwan.950  LC_CTYPE=Chinese (Traditional)_Taiwan.950
# [3] LC_MONETARY=Chinese (Traditional)_Taiwan.950 LC_NUMERIC=C
# [5] LC_TIME=Chinese (Traditional)_Taiwan.950
#
# attached base packages:
# [1] stats     graphics  grDevices utils     datasets  methods   base
#
# other attached packages:
# 1] tidyr_0.5.1.9000    data.table_1.9.7    dplyr_0.5.0.9000    dtplyr_0.0.1.9000
# [5] plyr_1.8.4.9000     pipeR_0.6.1.3       RevoUtilsMath_3.2.5
#
# loaded via a namespace (and not attached):
# [1] Rcpp_0.12.5.2        assertthat_0.1.0.99  digest_0.6.9         withr_1.0.1
# [5] chron_2.3-47         R6_2.1.2             DBI_0.4-1            magrittr_1.5
# [9] git2r_0.14.0         httr_1.1.0           stringi_1.0-1        curl_0.9.7
# [13] lazyeval_0.1.10.9000 devtools_1.11.1      tools_3.2.5          memoise_1.0.0
# [17] knitr_1.12.3         tibble_1.0