R/randomizeCNOlist.R

In CellNOptR: Training of boolean logic models of signalling networks using prior knowledge networks and perturbation data

#
#  This file is part of the CNO software
#
#  Copyright (c) 2011-2013 - EBI
#
#  File author(s): CNO developers (cno-dev@ebi.ac.uk)
#
#  Distributed under the GPLv3 License.
#  See accompanying file LICENSE.txt or copy at
#      http://www.gnu.org/licenses/gpl-3.0.html
#
#  CNO website: http://www.cellnopt.org
#
##############################################################################
# $Id$

# randomize a cnolist by adding noise to the signal matrices at all times except time=0
# The random noise is a gaussian distribution, which standard deviation can be
# changed.
# The random noise can also be a uniform noise between 0 and 1, in which case
# the data is not used anymore.
randomizeCNOlist <- function(cnolist, sd=0.1, minValue=0, maxValue=1, mode="gaussian"){

    if (mode %in% list("gaussian", "uniform", "shuffle")==FALSE){
        stop("mode can be only 'gaussian', 'uniform', or 'shuffle'")
    }

    cnames = colnames(cnolist@signals$`0`)
    dimensions = dim(cnolist@signals$`0`)
    times = cnolist@timepoints


    for (time in 2:length(times)){
        if (mode=="gaussian"){
            cnolist@signals[[time]] = rnorm(cnolist@signals[[time]], sd=sd)
        }
        if (mode=="uniform"){
            cnolist@signals[[time]] = runif(cnolist@signals[[time]])
        }
        if (mode=="shuffle"){
            cnolist@signals[[time]] = sample(getSignals(c)[[time]])

        }
        # make sure that values are still in the expected range
        cnolist@signals[[time]][cnolist@signals[[time]]<minValue] <- minValue
        cnolist@signals[[time]][cnolist@signals[[time]]>maxValue] <- maxValue

        # set back the column names that are lost
        dim(cnolist@signals[[time]])<-dimensions
        colnames(cnolist@signals[[time]])<-cnames
    }

     return(cnolist)
 }