adam_cowlog.R

require(tidyverse)

# computes things you want to know across the entire trial
do_it <- function(x) {
  name <- x$filename[1]
  fish_name <- name %>% stringr::str_match(pattern = "_S[a-z]+?_([A-Za-z]+)_") %>% as.vector %>% .[2]
  
  temp <- x %>% 
  mutate(time_zone = lead(time) - time) %>% 
  group_by(code) %>%
  summarise(total = sum(time_zone, na.rm = TRUE)) %>% 
  spread(code, total)
  
  temp$transits <- nrow(x)
  temp$filename <- name
  temp$name = stringr::str_replace_all(string = fish_name, pattern ="[//_]", "")
  temp
}

# computes things you want to know for each half
do_it_with_halves <- function(x) {
  name <- x$filename[1]
  fish_name <- name %>% stringr::str_match(pattern = "_S[a-z]+?_([A-Za-z]+)_") %>% as.vector %>% .[2]
  
  total_time <- x$time[length(x$time)]
  half_time <- round(total_time / 2, 1)
  
  if(!half_time %in% x$time){
    # find the right index to add it to the dataframe
    idx <- max(which(x$time < half_time))
    x <- add_row(x, time = half_time, class = 1, code = x$code[idx], .after = idx) %>%
      mutate(half = if_else(time < half_time, "first_half", "second_half"))
  }
  
  temp <- x %>% 
    mutate(time_zone = lead(time) - time) %>% 
    group_by(code, half) %>%
    summarise(total = sum(time_zone, na.rm = TRUE)) %>% 
    spread(code, total)
  
  temp$transits <- c(filter(x, half == "first_half") %>% nrow, filter(x, half == "second_half") %>% nrow)
  temp$filename <- name
  temp$name = stringr::str_replace_all(string = fish_name, pattern ="[//_]", "")
  temp
}


path <- "path/to/your/cowlog/files"
scototaxis <- dir(path, pattern = "*.csv", full.names = TRUE) %>%
  set_names %>%
  map(read_csv) %>%
  imap(~ transform(.x, filename = .y)) %>%
  map(do_it_with_halves) %>%
  Reduce(bind_rows, .)