R/get_results_daj.R

In antonmalko/ibextor: Easy preprocessing of Ibex data

get_results_daj <- function(file_name,
                           elem_number = NULL,
                           del_col = NULL, del_mode = "auto",
                           col_names = NULL,
                           partial_names = TRUE,
                           col_classes = NULL, partial_classes = TRUE,
                           short_subj_ids = TRUE,
                           sprt = TRUE,
                           ...) {

  #' @rdname get_results
  #' @export


  res <- read_ibex(file_name, ...)
  res <- subset_ibex(res, controller = "DashedAcceptabilityJudgment",
                     elem_number = elem_number)

  sprt_lines <- detect_sprt_lines(res)

  if (all(sprt_lines) == TRUE & !sprt){ # if the auto detection says it sprt, but the user requested speeded acceptability
    stop("Column 10 doesn't contain numeric data or there are more than 11 columns in the data.
         It is possible that you are subsetting for the wrong controller, or that the mode of ",
         "DashedSentence controller was set to `self-paced reading` during ",
         "data collection, while you are requesting `acceptability judgment` (sprt = FALSE)")
  } else {
    if (!any(sprt_lines) == TRUE & sprt) { # if the auto detection says it speeded acceptability, but the user requested sprt
      stop("Column 8 contains non-numeric data, which it shouldn't. It is possible ",
           "that you are subsetting for the wrong controller, or that the mode of ",
           "DashedSentence controller was set to `acceptability judgment` during ",
           "data collection, while you are requesting `self-paced reading` (sprt = TRUE)")
    } else { # if the data is a mixture of both, only pick one subset
      warning("The data seems to contain a mixture of sentences in `self-paced reading` and ",
              "`speeded acceptability` mode. Will only choose sentence for ",
              ifelse(sprt, "self-paced reading", "speeded acceptability"), " mode",
              " since sprt = ", sprt)
    }
  }



  # --------- 1. Speeded acceptability mode ----------------------
  if (sprt){

    # if the mode is sprt, just add question lines to sprt lines and extract them
    extract_lines <- add_quest_lines(sprt_lines)
    res <- res[extract_lines,]

    if (is.null(col_names)){
      col_names <- c("region",
                     "word",
                     "rt",
                     "line_break",
                     "sentence",
                     "question",
                     "answer",
                     "is_correct",
                     "quest_rt")
    }

    if (is.null(col_classes)){
      col_classes <- c("factor",
                       "character",
                       "numeric",
                       "logical",
                       "character",
                       "character",
                       "character",
                       "numeric",
                       "numeric")
    }

    # Subset for sentences
    res_sent <- res[res[,12]!="",  ]
    if (NROW(res_sent)==0) stop ("Subsetting for sentence data failed")
    rownames(res_sent) <- NULL
    res_sent <- droplevels(res_sent)

    # Subset for Questions. Get only the relevant columns, i.e 8-11
    res_quest <- res[res[,12]=="", c(4, 8:11)]
    if (NROW (res_quest)==0) stop ("Failed to get questions data")
    rownames(res_quest) <- NULL

    # merge on the 4th column, which is item number.
    # The subsetting at the end prevent "merge" from reordering the columns,
    # otherwise, it puts the "by" column first, which messes up later
    # column naming, class assignment and column deletion
    res <- merge(res_sent, res_quest, by.x = 4, by.y = 1, sort = FALSE)
    res <- res[, c(2:4, 1, 5:ncol(res))]

    res <- format_ibex(res,
                       col_names = col_names, partial_names = partial_names,
                       col_classes = col_classes, partial_classes = partial_classes)
    res <- recode_subjects(res, short_ids = short_subj_ids)
    res <- delete_columns(res, del_col, del_mode)
    return(res)


    } else {

    sa_lines <- detect_sa_lines(res)
    extract_lines <- add_quest_lines(sa_lines)
    res <- res[extract_lines,]

    if (is.null(col_names)){
      col_names <- c("question",
                     "answer",
                     "is_correct",
                     "quest_rt",
                     "sentence")
    }

    if (is.null(col_classes)){
      col_classes <- c("character",
                       "character",
                       "logical",
                       "numeric",
                       "character")
    }

    res <- res[,1:11]

    #------------ read the data ------------------------

    # read odd lines with the info about Sentences. Here we read only the
    # last column, which has number 8
    res_sent <- res[(seq(1, NROW(res), by=2)), 8 ]
    if (NROW(res_sent)==0) stop ("Subsetting for sentence data failed")


    # read even lines, which contain info about questions.
    # we read all the columns, so almost all the information is already extracted
    res <- res[(seq(2, NROW(res), by=2)),  ]
    if (NROW(res)==0) stop ("Subsetting for questions data failed")
    rownames(res) <- NULL

    # add column with question info after immediately after all the other info,
    res[,12] <- res_sent
    droplevels(res)

    res <- format_ibex(res,
                       col_names = col_names, partial_names = partial_names,
                       col_classes = col_classes, partial_classes = partial_classes)
    res <- recode_subjects(res, short_ids = short_subj_ids)
    res <- delete_columns(res, del_col, del_mode)

    return (res)
    }
}