\name{readMgfData}
\alias{readMgfData}

\title{
  Import mgf files as 'MSnExp' instances.
}

\description{
  Reads an mgf file and generates an \code{"\linkS4class{MSnExp}"}
  object. 
}

\usage{
readMgfData(file, pdata = NULL, centroided = TRUE, smoothed = FALSE,
verbose = TRUE, cache = 1)
}


\arguments{
  \item{file}{
    character vector with file name to be read.
  }
  \item{pdata}{
    an object of class \code{"\linkS4class{NAnnotatedDataFrame}"}.
  }
  \item{smoothed}{
    \code{Logical} indicating whether spectra already smoothed or not. Default is
    'FALSE'. Used to initialise \code{"\linkS4class{MSnProcess}"} object
    in \code{processingData} slot.
  }
  \item{centroided}{
    \code{Logical} indicating whether spectra are centroided or not. Default is
    'TRUE'. Used to initialise \code{"\linkS4class{MSnProcess}"} object
    in \code{processingData} slot.
  }
  \item{cache}{
    Numeric indicating caching level. Default is 1. Under development.
  }
  \item{verbose}{
    verbosity flag. 
  }
}

\value{
  An instance of 
}

\author{
  Guangchuang Yu <guangchuangyu@gmail.com> and Laurent Gatto <lg390@cam.ac.uk>
}

\details{
  Note that when reading an mgf file, the original order of the spectra
  is lost. Thus, if the data was originally written to mgf from an
  \code{MSnExp} object using \code{writeMgfData}, although the feature
  names will be identical, the spectra are not as a result of the
  reordering. See example below.
}

\seealso{
  \code{\link{writeMgfData}} method to write the content of
  \code{"\linkS4class{Spectrum}"} or \code{"\linkS4class{MSnExp}"}
  objects to mgf files. Raw data files can also be read with the
  \code{\link{readMSData}} function.
  }
  
\examples{
  \dontrun{
    data(itraqdata)
    writeMgfData(itraqdata,file="itraqdata.mgf",COM="MSnbase itraqdata")
    itraqdata2 <- readMgfData("itraqdata.mgf")
    ## note that the order of the spectra is altered
    ## and precision of some values (precursorMz for instance)
    match(signif(precursorMz(itraqdata2),4),signif(precursorMz(itraqdata),4))
    ## [1]  1 10 11 12 13 14 15 16 17 18 ...
    ## ... but all the precursors are there
    all.equal(sort(precursorMz(itraqdata2)),sort(precursorMz(itraqdata)),
              check.attributes=FALSE,
              tolerance=10e-5)
    ## is TRUE
    all.equal(as.data.frame(itraqdata2[[1]]),as.data.frame(itraqdata[[1]]))
    ## is TRUE
    all.equal(as.data.frame(itraqdata2[[3]]),as.data.frame(itraqdata[[11]]))
    ## is TRUE
  }
}

\keyword{ file }