学習者コーパス論 2015.11.13 宿題

LC-2015.11.13.R

library(magrittr)
library(stringr)

# read in all the files
files_nns <- list.files('NICE3.0/NICE-NNS', full = TRUE)
files_ns  <- list.files('NICE3.0/NICE-NS', full = TRUE)

nns <- lapply(files_nns, scan, what = 'char', sep = '\n', fileEnc = 'cp932')
ns  <- lapply(files_ns,  scan, what = 'char', sep = '\n', fileEnc = 'cp932')


# extract the data
nns_essay <- do.call(c, nns) %>%
    str_extract('(?<=^\\*JPN\\d{3}:\\t).*') %>%
    na.omit()

ns_essay  <- do.call(c, ns) %>%
    str_extract('(?<=^\\*NS\\d{3}:\\t).*') %>%
    na.omit()

# combine the two data
essays <- list(NNS = nns_essay, NS = ns_essay)

# count of words
count_words <- sapply(essays, function(x) {
    x %>%
    str_count('\\S+') %>%
    sum()
})

# count of genitives
count_of <- sapply(essays, function(x) {
    x %>% 
    str_count('\\b[Oo]f\\b') %>%
    sum()
})

count_s <- sapply(essays, function(x) {
    x %>% 
    str_count("('s|s')\\b") %>%
    sum()
})


# result
result <- 
    data.frame(
        words = count_words,
        of    = count_of,
        s     = count_s) %>%
    transform(
        gen_per_word = (of + s) / words,
        s_per_gen = s / (of + s)
        )

round(result, 3)