R/AllClasses.R

In thomazbastiaanssen/anansi: Annotation-Based Analysis of Specific Interactions

#' An S4 class to contain all metabolomics and functional input data as well as a dictionary to link them.
#'
#' @slot tableY A matrix of metabolomics data. Rows are samples and columns are features.
#' @slot tableX A matrix of functional data. Rows are samples and columns are features.
#' @slot dictionary A binary adjacency matrix. Typically generated using the \code{weaveWebFromTables()} function.
#' @description anansiWeb is the main container that will hold your input data thoughout the \code{anansi} pipeline.
#'
setClass("anansiWeb",
  slots = c(
    tableY     = "matrix",
    tableX     = "matrix",
    dictionary = "matrix"
  )
)


#' An S4 class to contain all \code{anansi} stats results so that they can easily be extracted.
#'
#' @slot subject A character that describes the data that was queried.
#' @slot type A character that describes type of parameter contained in the \code{estimates} slot.
#' For example r.values for correlations or r.squared for models.
#' @slot df a vector of length 2, containing df1 and df2 corresponding to the F-ratio considered.
#' @slot F.valules A matrix containing the f-values
#' @slot estimates A matrix containing the estimates for the parameters named in the \code{type} slot.
#' @slot p.values A matrix containing the p.values for the parameters named in the \code{type} slot.
#' @slot q.values A matrix containing the q.values for the parameters named in the \code{type} slot.
#' @description \code{anansiTale} is the main container that will hold your stats output data coming out of the \code{anansi} pipeline.
#'
setClass("anansiTale",
  slots = c(
    subject   = "character",
    type      = "character",
    df        = "numeric",
    estimates = "matrix",
    F.values  = "matrix",
    p.values  = "matrix",
    q.values  = "matrix"
  )
)



#' An S4 class to contain all \code{anansi} input so that they can easily be extracted.
#'
#' @slot input A list that holds the input data in \code{anansiWeb} format, as well as the \code{int.terms} and corresponding \code{groups}, the equivalent \code{lm.formula} and \code{error.term} argument(s) if provided.
#' @description \code{anansiInput} is the container that will hold your input data in the \code{anansiYarn} output file coming out of the \code{anansi} pipeline.
#'
setClass("anansiInput",
  slots = c(
    web = "anansiWeb",
    lm.formula = "formula",
    error.term = "ANY",
    int.terms = "vector",
    groups = "ANY"
  )
)

#' An S4 class to contain all \code{anansi} output so that they can easily be extracted.
#'
#' @slot outut A list that holds the output data in \code{anansiWeb} format, as well as the \code{groups} argument if provided.
#' @description \code{anansiOutput} is the container that will hold your output data in the \code{anansiYarn} output file coming out of the \code{anansi} pipeline.
#'
setClass("anansiOutput",
  slots = c(
    cor_results = "list",
    model_results = "list"
  )
)

#' An S4 class to contain all \code{anansi} input and output so that they can easily be extracted.
#'
#' @slot input A list that holds the input data in \code{anansiWeb} format, as well as the \code{groups} argument if provided.
#' @slot output A list that holds the output in \code{anansiTale} format.
#' @description \code{anansiYarn} is the main container that will hold your output data coming out of the \code{anansi} pipeline.
#' The \code{spinToWide()} and \code{spinToLong()} functions can be used to extract result tables in wide and long format, respectively.
#'
setClass("anansiYarn",
  slots = c(
    input  = "anansiInput",
    output = "anansiOutput"
  )
)

#' Plotting method for \code{anansiYarn} object.
#' @description Makes a lot of histograms, useful to eyeball the p-value distribution.
#' @param x An \code{anansiYarn} object.
#' @importFrom graphics hist par
#' @export
#'
setMethod("plot", "anansiYarn", function(x) {
  par(ask = TRUE)
  plotnames <- c(names(x@output@cor_results), names(x@output@model_results))
  plotlist <- list(unlist(x@output@cor_results), unlist(x@output@model_results))
  dictionary <- x@input@web@dictionary

  for (p in seq_len(length(plotnames))) {
    tale <- unlist(plotlist)[[plotnames[p]]]

    # figure out if we're plotting r.values or r.squared.
    if (tale@type == "r.values") {
      hist(c(tale@estimates[dictionary]),
        xlim = c(-1, 1), xlab = tale@type, main = paste(tale@type, "of", tale@subject)
      )
    }
    if (tale@type == "r.squared") {
      hist(c(tale@estimates[dictionary]),
        xlim = c(0, 1), xlab = tale@type, main = paste(tale@type, "of", tale@subject)
      )
    }
    # plot p.values
    hist(c(tale@p.values[dictionary]),
      xlim = c(0, 1), xlab = "p.values",
      main = paste("p.values of", tale@subject),
      breaks = 20
    )

    # plot q.values
    hist(c(tale@q.values[dictionary]),
      xlim = c(0, 1), xlab = "q.values",
      main = paste("q.values of", tale@subject),
      breaks = 20
    )
  }
  par(ask = FALSE)
})

#' Show method for \code{anansiYarn} object.
#' @description method to print \code{anansiYarn} object by calling \code{show}.
#' Since anansiYarn objects are typically huge and unwieldy, also gives some tips on how to handle it.
#' @param object An \code{anansiYarn} object.
#' @importFrom utils str
#' @importFrom methods show
#' @export
#'
setMethod("show", "anansiYarn", function(object) {
  str(object)
  cat("\nThis is an anansiYarn S4 object. They tend to be very large so we here's a summary instead.
      You can parse it to a data.frame using the spinToWide() & spinToLong() functions,
      or you can manually explore it by using the @ operator. ")
})

#' Print method for \code{anansiYarn} object.
#' @description method to print \code{anansiYarn} object by calling \code{print}.
#' Since anansiYarn objects are typically huge and unwieldy, also gives some tips on how to handle it.
#' @param x An \code{anansiYarn} object.
#' @importFrom utils str
#' @export
#'
setMethod("print", "anansiYarn", function(x) {
  str(x)
  cat("\nThis is an anansiYarn S4 object. They tend to be very large so we here's a summary instead.
      You can parse it to a data.frame using the spinToWide() & spinToLong() functions,
      or you can manually explore it by using the @ operator. ")
})

#' Summary method for \code{anansiYarn} object.
#' @description method to print \code{anansiYarn} object by calling \code{summary}.
#' Since anansiYarn objects are typically huge and unwieldy, also gives some tips on how to handle it.
#' @param object An \code{anansiYarn} object.
#' @importFrom utils str
#' @export
#'
setMethod("summary", "anansiYarn", function(object) {
  str(object)
  cat("\nThis is an anansiYarn S4 object. They tend to be very large so we here's a summary instead.
      You can parse it to a data.frame using the spinToWide() & spinToLong() functions,
      or you can manually explore it by using the @ operator. ")
})