as.ftmsData.R

#' Convert Data to peakData Class
#' 
#' Converts a list object or several data.frames of FT-MS data to an object of the class 'peakData'. Objects of the class 'peakData' are lists with three obligatory components \code{e_data}, \code{f_data}, and \code{e_meta}. 
#' 
#' @param e_data a \eqn{p \times n + 1} data.frame of expression data, where \eqn{p} is the number of observed peaks and \eqn{n} is the number of samples. Each row corresponds to data for each peak. One column specifying a unique identifier for each peak/mass (row) must be present.
#' @param f_data a data.frame with \eqn{n} rows. Each row corresponds to a sample with one column giving the unique sample identifiers found in e_data column names and other columns providing qualitative and/or quantitative traits of each sample.
#' @param e_meta a data.frame with \eqn{p} rows. Each row corresponds to a peak/mass with one column giving a unique peak/identifier (must be named the same as the column in \code{e_data}) and other columns giving meta information. At a minimum a column giving the mass of each peak and a column giving molecular formulae or columns giving elemental counts must be present.
#' @param edata_cname character string specifying the name of the column containing a unique identifier for each peak/mass in \code{e_data} and \code{e_meta}.
#' @param fdata_cname character string specifying the name of the column containing the sample identifiers in \code{f_data}.
#' @param mass_cname character string specifying the name of the column containing the peak/mass identifiers in \code{e_meta}. Note: this is often the same as \code{edata_cname} for cases where mass is used as a unique identifier.
#' @param ... further arguments
#'
#' @details Objects of class 'peakData' contain some attributes that are referenced by downstream functions. These attributes must be specified (or added using available functions) to reference downstream functions for: Kendrick plots, Van Krevelen plots, and functions involving databases.
#'
#' If your data contains information about isotopic peaks (e.g. C13), you should specify the attribute \code{isotopic_cname} which gives the column in \code{e_meta} that contains an indicator of yes/no for each peak. Additionally, you must specify the attribute \code{isotopic_notation} which is a character string indicating the value in column \code{isotopic_cname} which indicates that a peak is isotopic. 
#' Currently, any peaks that are isotopic are removed from the dataset, as available methods (e.g. Van Krevelen plot) are not applicable to these peaks.
#' 
#' Attributes giving general information about the data object:
#' \tabular{ll}{
#' data_scale \tab character string giving the scale that the data is on. Valid options include 'log2', 'log10', 'log' (for natural log), 'pres' (for 0/1 presence/absence data), and 'abundance'. Default value is 'abundance'. \cr
#' \tab \cr
#' instrument_type \tab character string giving the type of FT-MS instrument data was generated by. Valid options are: "12T" and "21T". Defaults to "12T". This information is used to determine appropriate plotting functions for Van Krevelen, Kendrick, etc. plots. \cr
#' }
#' Attributes giving extra information in \code{f_data}:
#' \tabular{ll}{
#' extraction_cname \tab character string specifying the name of the column, in \code{f_data}, containing information as to what extraction method was used for a sample. Only necessary if \code{e_data} contains samples from multiple extraction methods. \cr
#' }
#' Attributes giving extra information in \code{e_meta}: 
#' \tabular{ll}{
#' mass_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the mass information for each peak. \cr
#' \tab \cr
#' mf_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the mass (empirical) formula for a peak/mass. \cr
#' \tab \cr
#' element_col_names \tab named list of character strings specifying element/isotope column names, in \code{e_meta}, containing the respective count for each peak/mass. \cr
#' \tab \cr
#' isotopic_cname \tab character string specifying the name of the column, in \code{e_meta}, containing information about whether each peak is isotopic or not. \cr
#' \tab \cr
#' isotopic_notation \tab character string specifying the value used in column \code{isotopic_cname} which indicates that a peak is isotopic. \cr
#' \tab \cr
#' ratio_cnames \tab named list of character strings specifying the name of the column, in \code{e_meta}, containing the respective ratio of two elements or isotopes for each peak/mass. \cr
#' \tab \cr
#' kmass_cname \tab a possibly named character vector specifying the name of the columns, in \code{e_meta}, containing the Kendrick Mass for each peak/mass.  Names should be any of 'CH2', 'CO2', 'H2', 'H2O', 'CHO' and correspond to the base compounds used to calculate each of the Kendrick Masses \cr
#' \tab \cr
#' kdefect_cname \tab a possibly named character vector specifying the name of the column, in \code{e_meta}, containing the Kendrick Defect for each peak/mass. Names should be any of 'CH2', 'CO2', 'H2', 'H2O', 'CHO' and correspond to the base compounds used to calculate each of the Kendrick Masses\cr
#' \tab \cr
#' nosc_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the NOSC value for each peak/mass \cr
#' \tab \cr
#' gfe_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the Gibb's Free Energy value for each peak/mass \cr
#' \tab \cr
#' mfname_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the name/description for each peak/mass \cr
#' \tab \cr
#' aroma_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the aromaticity value for each peak/mass \cr
#' \tab \cr
#' modaroma_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the modified aromaticity value for each peak/mass \cr
#' \tab \cr
#' dbe_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the double-bond equivalent values for each peak/mass \cr
#' \tab \cr
#' dbeo_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the double-bond equivalent minus oxygen value for each peak/mass \cr
#' \tab \cr
#' dbeai_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the double-bond equivalent aromaticity index value for each peak/mass \cr
#' \tab \cr
#' elcomp_cname \tab character string specifying the name of the column, in \code{e_meta}, containing the general elemental composition of each peak/mass \cr
#' \tab \cr
#' check_rows \tab logical indicating whether to remove peaks with no nonzero entries.  Defaults to FALSE \cr
#' }
#'
#'@author Lisa Bramer
#'
#'@export

as.peakData <- function(e_data, f_data, e_meta, edata_cname, fdata_cname, mass_cname, ...){
  .as.peakData(e_data, f_data, e_meta, edata_cname, fdata_cname, mass_cname, ...)
}

# Need to change the col names param here 
.as.peakData <- function(e_data, f_data, e_meta, edata_cname, fdata_cname, mass_cname,
                        extraction_cname = NULL, mf_cname = NULL, 
                        element_col_names = list("C"="C", "H"="H", "O"=NULL, "N"=NULL, "S"=NULL, "P"=NULL),
                        isotopic_cname = NULL, isotopic_notation = NULL, 
                        ratio_cnames = list("O:C" = NULL, "H:C" = NULL), kmass_cname = NULL,
                        kdefect_cname = NULL, nosc_cname = NULL, gfe_cname = NULL, mfname_cname = NULL, 
                        aroma_cname = NULL, modaroma_cname = NULL, dbe_cname = NULL, dbeo_cname = NULL, dbeai_cname = NULL,
                        elcomp_cname = NULL, instrument_type = "12T", data_scale = "abundance", check_rows = FALSE){
  

  # check that e_data, f_data, and e_meta are data.frames #
  # if(!inherits(e_data, "data.frame")) stop("e_data must be of the class 'data.frame'")
  # if(!inherits(f_data, "data.frame")) stop("f_data must be of the class 'data.frame'")
  # if(!inherits(e_meta, "data.frame")) stop("e_meta must be of the class 'data.frame'")
  
  # make sure e_data, f_data, and e_meta are data.frames #
  e_data <- as.data.frame(e_data)
  f_data <- as.data.frame(f_data)
  e_meta <- as.data.frame(e_meta)
  
  # check that the peak column exists in e_data and e_meta (if applicable) #
  if(!(edata_cname %in% names(e_data))) stop(paste("Peak/Mass column ", edata_cname," not found in e_data. See details of as.peakData for specifying column names.", sep = ""))
  if(!(edata_cname %in% names(e_meta))) stop(paste("Peak/Mass column ", edata_cname," not found in e_meta. Column names for peak/mass identifiers must match for e_data and e_meta. See details of as.peakData for specifying column names.", sep = ""))
  
  # check that e_data has unique rows #
  if(nrow(e_data) != length(unique(dplyr::pull(e_data, edata_cname)))) stop("The 'edata_cname' identifier is non-unique.")

  # check that f_data has unique rows #
  if(nrow(f_data) != length(unique(dplyr::pull(f_data, fdata_cname)))) stop("The 'fdata_cname' identifier is non-unique.")

  # check that instrument_type is a valid string #
  if(!(instrument_type %in% c("21T", "12T"))) stop("Instrument type is not valid. See details of as.peakData for valid options.")

  # check that mass_cname is not null #
  if(is.null(mass_cname)) stop("'mass_cname' must be specified")
  
  # check that either mf_cname or elemental cnames are not null #
  if(is.null(mf_cname) & any(c(is.null(element_col_names$C), is.null(element_col_names$H)))) {
    stop("Either 'mf_cname' or both 'C' and 'H' must be specified in 'element_col_names'")
  }
  # check that cname arguments are found #
  if(!(mass_cname %in% names(e_meta))) {
    stop(paste("Mass column", mass_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = ""))
  }
  if(!is.null(element_col_names$C)){
    if(!(element_col_names$C %in% names(e_meta))) stop(paste("Carbon column", element_col_names$C, " not found in e_meta. See details of as.peakData for specifying column names.", sep = ""))
  }
  if(!is.null(element_col_names$H)){
    if(!(element_col_names$H %in% names(e_meta))) stop(paste("Hydrogen column", element_col_names$H, " not found in e_meta. See details of as.peakData for specifying column names.", sep = ""))
  }
  # Check that the elements/isotopes in element_col-names are within coreMS element/isotope superset.
  if(!all(names(element_col_names) %in% element_names)){
    stop(paste("The following elements in 'element_col_names' are not supported: ", paste(names(element_col_names[!(names(element_col_names) %in% element_names)]), collapse = ','), ". See details of as.peakData for specifying column names.", sep = ""))
  }
  # Check that non-NULL elements in element_col_names are present within e_meta
  for(element in names(element_col_names)){
    if(!is.null(element_col_names[[element]]) && ! element %in% names(e_meta)){
      stop(paste("The following element in 'element_col_names' is not found in e_meta: ", element))
    }
  }
  if(!is.null(isotopic_cname)){
    if(!(isotopic_cname %in% names(e_meta))) stop(paste("Isotopic column", isotopic_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = ""))
  }
  if(!is.null(ratio_cnames$o2c_cname)){
    if(!(ratio_cnames$o2c_cname %in% names(e_meta))) stop(paste("Oxygen:Carbon Ratio column ", ratio_cnames$o2c_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(ratio_cnames$h2c_cname)){
    if(!(ratio_cnames$h2c_cname %in% names(e_meta))) stop(paste("Hydrogen:Carbon Ratio column ", ratio_cnames$h2c_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(kmass_cname)){
    if(!is.null(names(kmass_cname))){
      if(!all(names(kmass_cname)) %in% c('CH2', 'CO2', 'H2', 'H2O', 'CHO', '')) stop("Names (not the values) of kmass_cname must be in 'CH2', 'CO2', 'H2', 'H2O', 'CHO'")
    }
    if(!all(kmass_cname %in% names(e_meta))) stop(paste("Kendrick Mass column ", kmass_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(kdefect_cname)){
    if(!is.null(names(kdefect_cname))){
      if(!all(names(kdefect_cname)) %in% c('CH2', 'CO2', 'H2', 'H2O', 'CHO', '')) stop("Names (not the values) of kdefect_cname must be in 'CH2', 'CO2', 'H2', 'H2O', 'CHO'")
    }
    if(!all(kdefect_cname %in% names(e_meta))) stop(paste("Kendrick Defect column ", kdefect_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(nosc_cname)){
    if(!(nosc_cname %in% names(e_meta))) stop(paste("NOSC column ", nosc_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(gfe_cname)){
    if(!(gfe_cname %in% names(e_meta))) stop(paste("Gibbs Free Energy column ", gfe_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(mfname_cname)){
    if(!(mfname_cname %in% names(e_meta))) stop(paste("Molecular formula name/description column ", mfname_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(aroma_cname)){
    if(!(aroma_cname %in% names(e_meta))) stop(paste("Aromaticity column ", aroma_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(modaroma_cname)){
    if(!(modaroma_cname %in% names(e_meta))) stop(paste("Modified aromaticity column ", modaroma_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(mf_cname)){
    if(!(mf_cname %in% names(e_meta))) stop(paste("Molecular formula column ", mf_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(dbe_cname)){
    if(!(dbe_cname %in% names(e_meta))) stop(paste("Double-bond equivalent column ", dbe_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(dbeo_cname)){
    if(!(dbeo_cname %in% names(e_meta))) stop(paste("Double-bond equivalent (minus Oxygen) column ", dbeo_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(dbeai_cname)){
    if(!(dbeai_cname %in% names(e_meta))) stop(paste("Double-bond equivalent (aromaticity index) column ", dbeai_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(elcomp_cname)){
    if(!(elcomp_cname %in% names(e_meta))) stop(paste("Elemental composition column ", elcomp_cname, " not found in e_meta. See details of as.peakData for specifying column names.", sep = "") )
  }
  if(!is.null(extraction_cname)){
    if(!(extraction_cname %in% names(f_data))) stop(paste("Extraction Method column ", extraction_cname, " not found in f_data. See details of as.peakData for specifying column names.", sep = "") )
  }
  
  
  # check that the Sample column name is in f_data column names #
  if(!(fdata_cname %in% names(f_data))) stop(paste("Sample column ", fdata_cname, " not found in f_data. See details of as.pepData for specifying column names.", sep = ""))
  
  # check that all samples in e_data are present in f_data #
  edat_sampid = which(names(e_data) == edata_cname) 
  samps.miss = sum(!(names(e_data[,-edat_sampid]) %in% dplyr::pull(f_data,fdata_cname)))
  if( samps.miss > 0) stop(paste( samps.miss, " samples from e_data not found in f_data", sep = ""))
  
  # check for any extra samples in f_data not in e_data - necessary to remove before group_designation function #
  if(any(!(dplyr::pull(f_data, fdata_cname) %in% names(e_data)))){
    f_data <- f_data[-which(!(dplyr::pull(f_data,fdata_cname) %in% names(e_data))),]
    warning("Extra samples were found in f_data that were not in e_data. These have been removed from f_data.")
  }
  
  # check that f_data has at least 2 columns #
  # if(ncol(f_data) < 2) stop("f_data must contain at least 2 columns")
  
  # if e_meta is provided, check that all peaks in e_data occur in e_meta #
  if(!is.null(e_meta)){
    if(sum(!(dplyr::pull(e_data,edata_cname) %in% dplyr::pull(e_meta,edata_cname))) > 0 ) stop("Not all peaks in e_data are present in e_meta")
  }
  
  # if e_meta is provided, remove any extra features that were provided #
  if(!is.null(e_meta)){
    if(any(!(dplyr::pull(e_meta,edata_cname) %in% dplyr::pull(e_data,edata_cname)))){
      e_meta <- e_meta[-which(!(dplyr::pull(e_meta, edata_cname) %in% dplyr::pull(e_data, edata_cname))),]
      warning("Extra peaks were found in e_meta that were not in e_data. These have been removed from e_meta.")
    }
  }
  
  # convert the unique identifier column to character, to avoid factors #
  e_data[, edata_cname] = as.character(dplyr::pull(e_data, edata_cname))
  e_meta[, edata_cname] = as.character(dplyr::pull(e_meta, edata_cname))
  
  # check that if isotopic_cname is non-NULL then isotopic_notation is also non-NULL # 
  if(!is.null(isotopic_cname)){
    if(is.null(isotopic_notation)) stop("You must specify 'isotopic_notation', if 'isotopic_cname' is non-NULL")
  }
  
  # if isotopic_cname is not NULL #
  if(!is.null(isotopic_cname)){
    # filter out peaks where isotopic_cname = TRUE #
    ids_rmv = e_meta[which(as.character(e_meta[,isotopic_cname]) == isotopic_notation),edata_cname]
    if(length(ids_rmv) > 0){
      e_meta = e_meta[-which(dplyr::pull(e_meta, edata_cname) %in% ids_rmv), ]
      e_data = e_data[-which(dplyr::pull(e_data, edata_cname) %in% ids_rmv), ]
    }
  }
  
  # if mf_cname is NULL and any of o_cname, n_cname, s_cname or p_cname are NULL, create columns of zeroes for them #
  if (is.null(mf_cname) & is.null(element_col_names$O)) {
    element_col_names$O <- tail(make.unique(c(colnames(e_meta), "O")), 1) #unique column name
    e_meta[[element_col_names$O]] <- 0
  }
  if (is.null(mf_cname) & is.null(element_col_names$N)) {
    element_col_names$N <- tail(make.unique(c(colnames(e_meta), "N")), 1) #unique column name
    e_meta[[element_col_names$N]] <- 0
  }
  if (is.null(mf_cname) & is.null(element_col_names$S)) {
    element_col_names$S <- tail(make.unique(c(colnames(e_meta), "S")), 1) #unique column name
    e_meta[[element_col_names$S]] <- 0
  }
  if (is.null(mf_cname) & is.null(element_col_names$P)) {
    element_col_names$P <- tail(make.unique(c(colnames(e_meta), "P")), 1) #unique column name
    e_meta[[element_col_names$P]] <- 0
  }
  
  # store results #
  res = list(e_data = e_data, f_data = f_data, e_meta = e_meta)
  
  # set column name attributes #
  attr(res, "cnames") = list(edata_cname = edata_cname, fdata_cname = fdata_cname, mass_cname = mass_cname, 
                             extraction_cname = extraction_cname, mf_cname = mf_cname, 
                             element_col_names = element_col_names, isotopic_cname = isotopic_cname,
                             ratio_cnames = ratio_cnames, kmass_cname = kmass_cname, kdefect_cname = kdefect_cname, 
                             nosc_cname = nosc_cname, gfe_cname = gfe_cname, mfname_cname = mfname_cname, 
                             aroma_cname = aroma_cname, modaroma_cname = modaroma_cname, dbe_cname = dbe_cname,
                             dbeo_cname = dbeo_cname, dbeai_cname = dbeai_cname, elcomp_cname = elcomp_cname 
                             )
  
  attr(res, "data_info") = list(data_scale = data_scale, instrument_type = instrument_type)
  
  # set group dataframe attribute to NULL, will be filled in after running group_designation function #
  attr(res, "group_DF") = NULL

  # set class of list #
  class(res) = c("peakData","ftmsData")
  
  # check for empty rows and remove them with molfilt 
  if(check_rows){
    molfilt <- molecule_filter(res)
    if(any(molfilt$Num_Observations == 0)) res <- applyFilt(molfilt, res, min_num = 1)
  }
  
  # set filters attributes #
  attr(res, "filters") = NULL
  
  #if mf_cname is NULL and elemental columns are non-NULL, construct formulae #
  if(is.null(mf_cname) & all(!sapply(element_col_names, is.null))){
    res = assign_mf(res)
  }

  # # if mf_cname is non-NULL and any elements within element_cols are NULL, parse formulae #
  if(!is.null(mf_cname) & any(sapply(element_col_names, is.null))){
    res = parse_mf(res)
  }
  
  return(res)
  
}


#' Convert Data to compoundData Class
#' 
#' Converts a list object or several data.frames of FT-MS data to an object of the class 'compoundData'. Objects of the class 'compoundData' are lists with three obligatory components \code{e_data}, \code{f_data}, and \code{e_meta}. 
#' 
#' @param e_data a \eqn{p \times n + 1} data.frame of expression data, where \eqn{p} is the number of observed compounds and \eqn{n} is the number of samples. Each row corresponds to data for each peak. One column specifying a unique identifier for each peak/mass (row) must be present.
#' @param f_data a data.frame with \eqn{n} rows. Each row corresponds to a sample with one column giving the unique sample identifiers found in e_data column names and other columns providing qualitative and/or quantitative traits of each sample.
#' @param e_meta a data.frame with \eqn{p} rows. Each row corresponds to a compound with one column giving a unique peak/identifier (must be named the same as the column in \code{e_data}) and other columns giving meta information. At a minimum a column giving the mass of each peak and a column giving molecular formulae or columns giving elemental counts must be present.
#' @param edata_cname character string specifying the name of the column containing a unique identifier for each peak/mass in \code{e_data} and \code{e_meta}.
#' @param fdata_cname character string specifying the name of the column containing the sample identifiers in \code{f_data}.
#' @param mass_cname character string specifying the name of the column containing the peak/mass identifiers in \code{e_meta}. Note: this is often the same as \code{edata_cname} for cases where mass is used as a unique identifier.
#' @param compound_cname character string specifying the name of the column containing the compound identifier in \code{e_meta}. This is a compound identifier related to a database (e.g. MetaCyc)
#' @param ... further arguments (see \code{\link{as.peakData}})
#'
#' @details \code{as.compoundData} constructs a compoundData object which is an ftmsData object where the rows of \code{e_data} correspond to compounds.
#' @rdname as.compoundData

as.compoundData <- function(e_data, f_data, e_meta, edata_cname, fdata_cname, mass_cname, compound_cname, ...){
  res <- as.peakData(e_data, f_data, e_meta, edata_cname, fdata_cname, mass_cname, ...)
  class(res) <- c("compoundData", "ftmsData")
  
  # if e_meta is provided, check that reaction_cname is in it
  if (!is.null(e_meta)) {
    if(!(compound_cname %in% names(e_meta))) stop(paste("Compound column", compound_cname, " not found in e_meta.", sep = ""))
  }
  res <- setCompoundColName(res, compound_cname)
  return(res)
}

#' Convert Data to reactionData Class
#' 
#' Converts a list object or several data.frames of FT-MS data to an object of the class 'reactionData'. Objects of the class 'reactionData' are lists with two obligatory components \code{e_data} and \code{f_data} and one optional compounent, \code{e_meta}. 
#' 
#' @param e_data a \eqn{p \times n + 1} data.frame of expression data, where \eqn{p} is the number of observed reactions and \eqn{n} is the number of samples. Each row corresponds to data for each peak. One column specifying a unique identifier for each peak/mass (row) must be present.
#' @param f_data a data.frame with \eqn{n} rows. Each row corresponds to a sample with one column giving the unique sample identifiers found in e_data column names and other columns providing qualitative and/or quantitative traits of each sample.
#' @param e_meta a data.frame with \eqn{p} rows. Each row corresponds to a reaction with one column giving a unique reaction identifier (must be named the same as the column in \code{e_data}) and other columns giving meta information. At a minimum a column giving the mass of each peak and a column giving molecular formulae or columns giving elemental counts must be present.
#' @param edata_cname character string specifying the name of the column containing a unique identifier for each reaction in \code{e_data} and \code{e_meta}.
#' @param fdata_cname character string specifying the name of the column containing the sample identifiers in \code{f_data}.
#' @param reaction_cname character string specifying the name of the column containing the reaction identifiers in \code{e_meta}.
#' @param instrument_type character string giving the type of FT-MS instrument data was generated by. Valid options are: "12T" and "21T". Defaults to "12T". This information is used to determine appropriate plotting functions for Van Krevelen, Kendrick, etc. plots.
#' @param db character string specifying the database from which reaction information is drawn
#' @param ... further arguments
#'
#' @details \code{as.reactionData} constructs a reactionData object which is an ftmsData object where the rows of \code{e_data} correspond to reactions.
#' @rdname as.reactionData
as.reactionData <- function(e_data, f_data, e_meta = NULL, edata_cname, fdata_cname, reaction_cname, instrument_type = "12T", db=NA, ...){
  
  # initial checks #
  
  # check that e_data, f_data, and e_meta are data.frames #
  if(!inherits(e_data, "data.frame")) stop("e_data must be of the class 'data.frame'")
  if(!inherits(f_data, "data.frame")) stop("f_data must be of the class 'data.frame'")
  if(!inherits(e_meta, "data.frame")) stop("e_meta must be of the class 'data.frame'")
  
  # check that the peak column exists in e_data and e_meta (if applicable) #
  if(!(edata_cname %in% names(e_data))) stop(paste("Column ", edata_cname," not found in e_data. See details of as.reactionData for specifying column names.", sep = ""))
  if(!(edata_cname %in% names(e_meta))) stop(paste("Column ", edata_cname," not found in e_meta. Column names for dentifiers must match for e_data and e_meta. See details of as.reactionData for specifying column names.", sep = ""))
  
  # check that e_data has unique rows #
  if(nrow(e_data) != length(unique(dplyr::pull(e_data, edata_cname)))) stop("The 'edata_cname' identifier is non-unique.")
  
  # check that f_data has unique rows #
  if(nrow(f_data) != length(unique(dplyr::pull(f_data, fdata_cname)))) stop("The 'fdata_cname' identifier is non-unique.")
  
  # check that the Sample column name is in f_data column names #
  if(!(fdata_cname %in% names(f_data))) stop(paste("Sample column ", fdata_cname, " not found in f_data. See details of as.pepData for specifying column names.", sep = ""))
  
  # check that all samples in e_data are present in f_data #
  edat_sampid = which(names(e_data) == edata_cname) 
  samps.miss = sum(!(names(e_data[,-edat_sampid]) %in% dplyr::pull(f_data, fdata_cname)))
  if( samps.miss > 0) stop(paste( samps.miss, " samples from e_data not found in f_data", sep = ""))
  
  # check for any extra samples in f_data not in e_data - necessary to remove before group_designation function #
  if(any(!(dplyr::pull(f_data, fdata_cname) %in% names(e_data)))){
    f_data <- f_data[-which(!(dplyr::pull(f_data, fdata_cname) %in% names(e_data))),]
    warning("Extra samples were found in f_data that were not in e_data. These have been removed from f_data.")
  }
  
  # check that f_data has at least 2 columns #
  if(ncol(f_data) < 2) stop("f_data must contain at least 2 columns")
  
  # if e_meta is provided, check that all peaks in e_data occur in e_meta #
  if(!is.null(e_meta)){
    if(sum(!(dplyr::pull(e_data, edata_cname) %in% dplyr::pull(e_meta, edata_cname))) > 0 ) stop("Not all identifiers in e_data are present in e_meta")
  }
  
  # if e_meta is provided, remove any extra features that were provided #
  if(!is.null(e_meta)){
    if(any(!(dplyr::pull(e_meta, edata_cname) %in% dplyr::pull(e_data, edata_cname)))){
      e_meta <- e_meta[-which(!(dplyr::pull(e_meta, edata_cname) %in% dplyr::pull(e_data, edata_cname))),]
      warning("Extra rows were found in e_meta that were not in e_data. These have been removed from e_meta.")
    }
  }
  
  # if e_meta is provided, check that reaction_cname is in it
  if (!is.null(e_meta)) {
    if(!(reaction_cname %in% names(e_meta))) stop(paste("Reaction column", reaction_cname, " not found in e_meta.", sep = ""))
  }
  
  # convert the unique identifier column to character, to avoid factors #
  e_data[, edata_cname] = as.character(dplyr::pull(e_data, edata_cname))
  e_meta[, edata_cname] = as.character(dplyr::pull(e_meta, edata_cname))
  
  # store results #
  res = list(e_data = e_data, f_data = f_data, e_meta = e_meta)
  
  # set column name attributes #
  attr(res, "cnames") = list(edata_cname = edata_cname, fdata_cname = fdata_cname, reaction_cname=reaction_cname)
  
  # set group dataframe attribute to NULL, will be filled in after running group_designation function #
  attr(res, "group_DF") = NULL
  
  # set filters attributes #
  attr(res, "filters") = NULL
  
  attr(res, "data_info") <- list(data_scale=NA, instrument_type=instrument_type)
  
  attr(res, "DB") <- db
  
  class(res) <- c("reactionData", "ftmsData")
  return(res)
}


#' Convert Data to moduleData Class
#' 
#' Converts a list object or several data.frames of FT-MS data to an object of the class 'moduleData'. Objects of the class 'moduleData' are lists with two obligatory components \code{e_data} and \code{f_data} and one optional component, \code{e_meta}. 
#' 
#' @param e_data a \eqn{p \times n + 1} data.frame of expression data, where \eqn{p} is the number of observed module nodes and \eqn{n} is the number of samples. Each row corresponds to data for each peak. One column specifying a unique identifier for each peak/mass (row) must be present.
#' @param f_data a data.frame with \eqn{n} rows. Each row corresponds to a sample with one column giving the unique sample identifiers found in e_data column names and other columns providing qualitative and/or quantitative traits of each sample.
#' @param e_meta a data.frame with \eqn{p} rows. Each row corresponds to a module node with one column giving a unique module node identifier (must be named the same as the column in \code{e_data}) and other columns giving meta information. At a minimum a column giving the mass of each peak and a column giving molecular formulae or columns giving elemental counts must be present.
#' @param edata_cname character string specifying the name of the column containing a unique identifier for each module node in \code{e_data} and \code{e_meta}.
#' @param fdata_cname character string specifying the name of the column containing the sample identifiers in \code{f_data}.
#' @param module_cname character string specifying the name of the column containing the module identifiers in \code{e_meta}.
#' @param module_node_cname character string specifying the name of the column containing the module node identifiers in \code{e_meta}
#' @param instrument_type character string giving the type of FT-MS instrument data was generated by. Valid options are: "12T" and "21T". Defaults to "12T". This information is used to determine appropriate plotting functions for Van Krevelen, Kendrick, etc. plots.
#' @param db character string specifying the database from which reaction information is drawn
#' @param ... further arguments
#'
#' @details \code{as.moduleData} constructs a moduleData object which is an ftmsData object where the rows of \code{e_data} correspond to unique module nodes.
#' @rdname as.moduleData
# @param node_label_cname character string specifying the name of the column containing the display name for each module node, in \code{e_meta}

as.moduleData <- function(e_data, f_data, e_meta = NULL, edata_cname, fdata_cname, module_cname, module_node_cname, 
                             instrument_type = "12T", db=NA, ...){
  
  # initial checks #
  
  # check that e_data, f_data, and e_meta are data.frames #
  if(!inherits(e_data, "data.frame")) stop("e_data must be of the class 'data.frame'")
  if(!inherits(f_data, "data.frame")) stop("f_data must be of the class 'data.frame'")
  if(!inherits(e_meta, "data.frame")) stop("e_meta must be of the class 'data.frame'")
  
  # check that the peak column exists in e_data and e_meta (if applicable) #
  if(!(edata_cname %in% names(e_data))) stop(paste("Column ", edata_cname," not found in e_data. See details of as.moduleData for specifying column names.", sep = ""))
  if(!(edata_cname %in% names(e_meta))) stop(paste("Column ", edata_cname," not found in e_meta. Column names for identifiers must match for e_data and e_meta. See details of as.moduleData for specifying column names.", sep = ""))
  
  # check that e_data has unique rows #
  if(nrow(e_data) != length(unique(dplyr::pull(e_data, edata_cname)))) stop("The 'edata_cname' identifier is non-unique.")
  
  # check that f_data has unique rows #
  if(nrow(f_data) != length(unique(dplyr::pull(f_data, fdata_cname)))) stop("The 'fdata_cname' identifier is non-unique.")
  
  # check that the Sample column name is in f_data column names #
  if(!(fdata_cname %in% names(f_data))) stop(paste("Sample column ", fdata_cname, " not found in f_data. See details of as.pepData for specifying column names.", sep = ""))
  
  # check that all samples in e_data are present in f_data #
  edat_sampid = which(names(e_data) == edata_cname) 
  samps.miss = sum(!(names(e_data[,-edat_sampid]) %in% dplyr::pull(f_data, fdata_cname)))
  if( samps.miss > 0) stop(paste( samps.miss, " samples from e_data not found in f_data", sep = ""))
  
  # check for any extra samples in f_data not in e_data - necessary to remove before group_designation function #
  if(any(!(dplyr::pull(f_data, fdata_cname) %in% names(e_data)))){
    f_data <- f_data[-which(!(dplyr::pull(f_data, fdata_cname) %in% names(e_data))),]
    warning("Extra samples were found in f_data that were not in e_data. These have been removed from f_data.")
  }
  
  # check that f_data has at least 2 columns #
  if(ncol(f_data) < 2) stop("f_data must contain at least 2 columns")
  
  # if e_meta is provided, check that all peaks in e_data occur in e_meta #
  if(!is.null(e_meta)){
    if(sum(!(dplyr::pull(e_data, edata_cname) %in% dplyr::pull(e_meta, edata_cname))) > 0 ) stop("Not all identifiers in e_data are present in e_meta")
  }
  
  # if e_meta is provided, remove any extra features that were provided #
  if(!is.null(e_meta)){
    if(any(!(dplyr::pull(e_meta, edata_cname) %in% dplyr::pull(e_data, edata_cname)))){
      e_meta <- e_meta[-which(!(dplyr::pull(e_meta, edata_cname) %in% dplyr::pull(e_data, edata_cname))),]
      warning("Extra rows were found in e_meta that were not in e_data. These have been removed from e_meta.")
    }
  }
  
  # if e_meta is provided, check that reaction_cname is in it
  if (!is.null(e_meta)) {
    if(!(module_cname %in% names(e_meta))) stop(paste("Module column", module_cname, " not found in e_meta.", sep = ""))
    if(!(module_node_cname %in% names(e_meta))) stop(paste("Module node column", module_node_cname, " not found in e_meta.", sep = ""))
    # if(!is.null(node_label_cname))
    #    if (!(node_label_cname %in% names(e_meta))) stop(paste("Node label column", node_label_cname, " not found in e_meta.", sep = ""))
  }
  
  # convert the unique identifier column to character, to avoid factors #
  e_data[, edata_cname] = as.character(dplyr::pull(e_data, edata_cname))
  e_meta[, edata_cname] = as.character(dplyr::pull(e_meta, edata_cname))
  
  # store results #
  res = list(e_data = e_data, f_data = f_data, e_meta = e_meta)
  
  # set column name attributes #
  attr(res, "cnames") <- list(edata_cname = edata_cname, fdata_cname = fdata_cname, module_cname=module_cname, module_node_cname=module_node_cname)

  # set group dataframe attribute to NULL, will be filled in after running group_designation function #
  attr(res, "group_DF") = NULL
  
  # set filters attributes #
  attr(res, "filters") = NULL
  
  attr(res, "data_info") <- list(data_scale=NA, instrument_type=instrument_type)

  attr(res, "DB") <- db
  
  class(res) <- c("moduleData", "ftmsData")
  return(res)
}