map values

library(data.table)
library(plyr)
library(dplyr)
library(tidyr)
library(magrittr)

# data generation
set.seed(100)
dt = data.table(t(replicate(100, sample(letters, 5)))) %>%
  setnames(paste0("id", 1:5))
mapTable = data.table(value = letters, score = sample(1:100, 26))

# solution
dt %>% add_rownames %>% tbl_dt %>% gather(id, value, -rowname) %>%
  mutate(score = mapvalues(value, mapTable$value, mapTable$score)) %>%
  group_by(rowname) %>% summarise(med_score = median(score))

# Source: local data table [100 x 2]
#
#    rowname med_score
# 1        1        73
# 2        2        40
# 3        3        54
# 4        4        42
# 5        5        43
# 6        6        58
# 7        7        56
# 8        8        38
# 9        9        58
# 10      10        58
# ..     ...       ...