http://blog.appsilondatascience.com/rstats/2017/03/02/r-fast-lookup.html

hash-comparison.R

library(tibble)
library(data.table)
library(dplyr)
library(hashmap)
library(microbenchmark)


set.seed(123)
nkeys <- 10 ** 7
keylen <- 8
keys <- unique(stringi::stri_rand_strings(nkeys, keylen))
values <- rnorm(length(keys))
rand_key <- sample(keys, 1)

system.time({ data_tbl <- tibble(keys, values) })
#  user  system elapsed 
# 0.024   0.000   0.023 

system.time({ data_table <- data.table(keys, values) })
#  user  system elapsed 
# 0.048   0.036   0.083 

data_table2 <- copy(data_table)
system.time({ setkey(data_table2, keys) })
#  user  system elapsed 
# 7.864   0.000   7.368 

# slow to construct (SEXPs copied to C++ object)
system.time({ hash <- hashmap(keys, values) })
#  user  system elapsed 
# 3.840   0.308   4.149 


# sanity check
data_tbl %>% filter(keys == rand_key)
# # A tibble: 1 × 2
#       keys    values
#      <chr>     <dbl>
# 1 dimWYkVf 0.3093082

data_table[keys == rand_key]
#        keys    values
# 1: dimWYkVf 0.3093082

data_table2[.(rand_key), nomatch = 0L]
#        keys    values
# 1: dimWYkVf 0.3093082

hash[[rand_key]]
# [1] 0.3093082


microbenchmark::microbenchmark(
    "dplyr" = data_tbl %>% filter(keys == rand_key),
    "data.table" = data_table[keys == rand_key],
    "indexed DT" = data_table2[.(rand_key), nomatch = 0L],
    "hashmap" = hash[[rand_key]],
    times = 200L
)
# Unit: microseconds
#        expr        min          lq         mean      median          uq         max neval
#       dplyr 129863.170 135231.3850 271200.45996 147699.5085 221820.6025 1905643.772   200
#  data.table    479.377    555.9585    850.10595    725.1985    903.3740    3861.167   200
#  indexed DT    493.498    589.2130    868.77630    743.1165    914.4945    3624.291   200
#     hashmap     16.820     37.5895     70.70894     75.1950     86.8735     331.077   200


sessionInfo()
# R Under development (unstable) (2017-02-21 r72242)
# Platform: x86_64-pc-linux-gnu (64-bit)
# Running under: Debian GNU/Linux 8 (jessie)
# 
# Matrix products: default
# BLAS: /usr/local/lib64/R/lib/libRblas.so
# LAPACK: /usr/local/lib64/R/lib/libRlapack.so
# 
# locale:
#  [1] LC_CTYPE=en_US.UTF-8       LC_NUMERIC=C               LC_TIME=en_US.UTF-8       
#  [4] LC_COLLATE=en_US.UTF-8     LC_MONETARY=en_US.UTF-8    LC_MESSAGES=en_US.UTF-8   
#  [7] LC_PAPER=en_US.UTF-8       LC_NAME=C                  LC_ADDRESS=C              
# [10] LC_TELEPHONE=C             LC_MEASUREMENT=en_US.UTF-8 LC_IDENTIFICATION=C       
# 
# attached base packages:
# [1] stats     graphics  grDevices utils     datasets  methods   base     
# 
# other attached packages:
# [1] microbenchmark_1.4-2.1 hashmap_0.2.0          dplyr_0.5.0           
# [4] data.table_1.10.0      tibble_1.2            
# 
# loaded via a namespace (and not attached):
#  [1] Rcpp_0.12.9.4    codetools_0.2-15 assertthat_0.1   plyr_1.8.4       grid_3.4.0      
#  [6] R6_2.2.0         gtable_0.2.0     DBI_0.5-1        magrittr_1.5     scales_0.4.1    
# [11] ggplot2_2.2.1    stringi_1.1.2    lazyeval_0.2.0   tools_3.4.0      munsell_0.4.3   
# [16] compiler_3.4.0   colorspace_1.2-7