R/NMscanTables.R

In NMdata: Preparation, Checking and Post-Processing Data for PK/PD Modeling

##' Find and read all output data tables in Nonmem run
##' @param file the Nonmem file to read (normally .mod or .lst)
##' @param details If TRUE, metadata is added to output. In this case,
##'     you get a list. Typically, this is mostly useful if
##'     programming up functions which behavior must depend on
##'     properties of the output.
##' @param quiet The default is to give some information along the way
##'     on what data is found. But consider setting this to TRUE for
##'     non-interactive use. Default can be configured using
##'     NMdataConf.
##' @param col.tableno Nonmem includes a counter of tables in the
##'     written data files. These are often not useful. However, if
##'     col.tableno is TRUE (not default), this will be carried forward
##'     and added as a column called NMREP. Even if NMREP is generated
##'     by NMscanTables, it is treated like any other table column in
##'     meta (?NMinfo) data.
##' @param col.nmrep col.nmrep If tables are repeated, include a
##'     counter? It does not relate to the order of the $TABLE
##'     statements but to cases where a $TABLE statement is run
##'     repeatedly. E.g., in combination with the SUBPROBLEMS feature
##'     in Nonmem, it is useful to keep track of the table
##'     (repetition) number. If col.nmrep is TRUE, this will be
##'     carried forward and added as a column called NMREP. This is
##'     default behavior when more than one $TABLE repetition is found
##'     in data. Set it to a different string to request the column
##'     with a different name. The argument is passed to NMscanTables.
##' @param col.id name of the subject ID column. Used for calculation
##'     of the number of subjects in each table.
##' @param as.fun The default is to return data as a data.frame. Pass
##'     a function (say tibble::as_tibble) in as.fun to convert to
##'     something else. If data.tables are wanted, use
##'     as.fun="data.table". The default can be configured using
##'     NMdataConf.
##' @param col.row The name of the row counter column. Optional and
##'     only used to check whether the row counter is in the data.
##' @param skip.absent Skip missing output table files with a warning?
##'     Default is FALSE in which case an error is thrown.
##' @param meta.only If TRUE, tables are not read, only a table is
##'     returned showing what tables were found and some available
##'     meta information. Notice, not all meta information (e.g.,
##'     dimensions) are available because the tables need to be read
##'     to derive that.
##' @return A list of all the tables as data.frames. If details=TRUE,
##'     this is in one element, called data, and meta is another
##'     element. If not, only the data is returned.
##' @examples
##' tabs1 <- NMscanTables(system.file("examples/nonmem/xgxr001.lst", package="NMdata"))
##' @family DataRead
##' @import data.table
##' @export

NMscanTables <- function(file,as.fun,quiet,col.nmrep=TRUE,col.tableno=FALSE,col.id="ID",col.row,details,skip.absent=FALSE,meta.only=FALSE){
    
#### Section start: Dummy variables, only not to get NOTE's in package checks ####

    DV <- NULL
    PRED <- NULL
    RES <- NULL
    WRES <- NULL
    has.col.id <- NULL
    has.col.row <- NULL
    filetype <- NULL
    maxLength <- NULL
    full.length <- NULL
    firstlastonly <- NULL
    firstonly <- NULL
    lastonly <- NULL
    level <- NULL
    name <- NULL
    nid <- NULL
    noheader <- NULL
    scope <- NULL

###  Section end: Dummy variables, only not to get NOTE's in pacakge checks ####
    
    if(missing(quiet)) quiet <- NULL
    quiet <- NMdataDecideOption("quiet",quiet)
    if(missing(as.fun)) as.fun <- NULL
    as.fun <- NMdataDecideOption("as.fun",as.fun)

    if(missing(col.row)||(!is.null(col.row)&&is.na(col.row))||(is.character(col.row)&&any(col.row==""))) {
        col.row <- NULL
    }    
    col.row <- NMdataDecideOption("col.row",col.row)

    ## if(!missing(details)) message("NMscanTables: argument details is deprecated.")
    args <- getArgs()
    deprecatedArg(oldarg="details",args=args)
    
    dir <- dirname(file)
    extract.info <- function(x,NAME,default){
        r1 <- regexpr(paste0(NAME," *= *[^ ]*"),x,ignore.case=TRUE)
        rm1 <- regmatches(x,r1)
        info <- sub(paste0(NAME," *= *"),"",rm1,ignore.case=TRUE)
        if(length(info)==0&&!missing(default)) {
            info <- default
        }
        info
    }
    
    lines.table <- NMreadSection(file,section="TABLE",keep.name=FALSE,
                                 keep.comments=FALSE,keep.empty=FALSE,
                                 as.one=FALSE, simplify=FALSE)
    if(length(lines.table)==0) {
        if(!quiet) messageWrap("No TABLE sections found in control stream. Returning NULL",
                               fun.msg=message)
        return(NULL)
    }
    
    tab.files <- lapply(lines.table,function(x) {
        tab <- data.table(file=filePathSimple(dir,extract.info(x,"FILE"))
                         ,name=extract.info(x,"FILE")
                         ,firstonly=any(grepl("FIRSTONLY|FIRSTRECORDONLY|FIRSTRECONLY",x))
                         ,lastonly=any(grepl("LASTONLY",x))
                         ,firstlastonly=any(grepl("FIRSTLASTONLY",x))
                         ,format=extract.info(x,"FORMAT",default=" ")
                         ,noheader=any(grepl("(NOHEADER|NOLABEL)",x)))
        tab
    })

    meta <- rbindlist(tab.files)
    ## seperator is not being used, so left out from metadata
    ##sep=ifelse(grepl(",",format),","," "),
    meta[,`:=`(nrow=NA_real_
              ,ncol=NA_real_
               )]

    meta.tabs.strange <- meta[(firstonly+lastonly+firstlastonly)>1]
    if(nrow(meta.tabs.strange)){
        warning("Table(s) seems to have more than one of the firstonly, lastonly and firstlastonly options. Does this make sense? Look at table(s): ",paste(meta.tabs.strange[,name],collapse=", "))
    }
    tables <- list()

    meta[,exists:=file.exists(file)]
    meta[,filetype:="output"]    
    meta[,file.mtime:=file.mtime(file)]

    if(meta.only){
        setcolorder(meta,intersect(c("source","name","nrow","ncol","nid","level","scope","has.col.row","has.col.id","full.length","filetype","format",
                                     "file.mtime","file"),colnames(meta)))
        return(as.fun(meta))
    }

######## read data
    for(I in 1:nrow(meta)){
        if(meta[I,!exists]) {
            if(skip.absent) {
                messageWrap(paste0("NMscanTables: File not found: ",meta[I,file],". Skipping."),fun.msg=message)
                next
            }
            stop(paste0("NMscanTables: File not found: ",meta[I,file],". Did you copy the lst file but forgot table file?"))
        }
        
        tables[[I]] <- NMreadTab(meta[I,file],quiet=TRUE,col.nmrep=col.nmrep,col.tableno=col.tableno,showProgress=FALSE,as.fun=identity,header=meta[I,!noheader],col.table.name=FALSE)
### to not include NMREP when counting columns
        ## dim.tmp <- dim(tables[[I]][,!colnames(tables[[I]])=="NMREP",with=FALSE])
        dim.tmp <- dim(tables[[I]])
        meta[I,nrow:=dim.tmp[1]]
        meta[I,ncol:=dim.tmp[2]]
        
        if(meta[I,noheader]) {
            messageWrap("Using NOHEADER and NOLABEL options in $TABLE is only experimentally supported in NMdata. Please double check the resuling column names. NMdata functions can handle the recurring headers in Nonmem tables so these $TABLE options should not be needed.",fun.msg=message)
            cnames.text <- paste(lines.table[[I]],collapse=" ")
            cnames.text <- gsub(","," ",cnames.text)
            cnames.text <- sub("^ *","",cnames.text)
            cnames.text <- sub(" +"," ",cnames.text)
            
            
            ## this is experimental. The list isn't exhaustive, and could these be column names if mixed in earlier in the $TABLE statement? Or are they all reserved names?
            cnames.notcols <- c("NOHEADER","FIRSTONLY","NOLABEL","ONEHEADER","LASTONLY","PRINT","NOPRINT","FILE *=")
            cnames.text <- sub(paste0("(",cnames.notcols,collapse="|",").*"),"",cnames.text)

            cnames.all <- strsplit(cnames.text," ")[[1]]
                        
            cnames.extra <- cc(DV,PRED,RES,WRES)
            cnames.extra <- setdiff(cnames.extra,cnames.all)
            ncol.I <- ncol(tables[[I]])
            ## if some were named by NMreadTable, we want to keep
            ## those names - they are most likely NMREP etc
            idx.named <- grep("V[1-9][0-9]*",colnames(tables[[I]]),invert=TRUE)
            if(length(idx.named)){
                ncol.I <- min(c(ncol.I,idx.named-1))
            }
            nce <- length(cnames.extra)
            
            if(nce){
                ## Adding 1 to indeces because NMreadTab adds NMREP
                cnames.all[(ncol.I-nce+1):(ncol.I)] <- cnames.extra
            }
            cnames <- cnames.all[1:ncol.I]
            
            setnames(tables[[I]],1:length(cnames),cnames)
        }
        
        cnames.tab.I <- colnames(tables[[I]])

        if(col.id%in%cnames.tab.I){
            meta[I,nid:=tables[[I]][,uniqueN(get(col.id))]]
        } else {
            meta[I,nid:=NA_real_]
        }
        
        if(!is.null(col.row)){
            meta[I,has.col.row:=col.row%in%cnames.tab.I]
        }
        if(!is.null(col.id)){
            meta[I,has.col.id:=col.id%in%cnames.tab.I]
        }
    }

    
    has.row.level <- meta[,any((firstonly+lastonly+firstlastonly)==0)]
    ## level is the observed level. If a first only table has as many
    ## rows as a row-level table, it's row label.
    if(has.row.level){
        meta[,full.length:=(nrow==max(nrow,na.rm=TRUE))]
        meta[full.length==TRUE,level:="row"]
        meta[full.length==FALSE,level:="id"]
    } else {
        meta[,full.length:=FALSE]
        meta[!is.na(nrow),level:="id"]
    }
    ## not supported
    meta[firstlastonly==TRUE,level:=NA_character_]
    
    meta[exists==TRUE,scope:="all"]
    meta[firstonly==TRUE,scope:="firstonly"]
    meta[lastonly==TRUE,scope:="lastonly"]
    meta[firstlastonly==TRUE,scope:="firstlastonly"]
    meta[,`:=`(firstonly=NULL,
               lastonly=NULL,
               firstlastonly=NULL
               )]


    ## sep not used so omitted 
    setcolorder(meta,intersect(c("source","name","nrow","ncol","nid","level","scope","has.col.row","has.col.id","full.length","filetype","format",
                                 "file.mtime","file"),colnames(meta)))
    
    
    if(!quiet){
        msg <- paste0("Number of output tables read: ",meta[exists==TRUE,.N])
        NIDL <- meta[exists==TRUE&level=="id",.N]
        NFLO <- meta[exists==TRUE&scope=="firstlastonly",.N]
        mNIDL <- NULL
        if(NIDL>0) mNIDL <- paste0(NIDL, " idlevel")
        mNFLO <- NULL
        if(NFLO>0) mNFLO <- paste0(NFLO, " FIRSTLASTONLY")
        msg.extra <- paste0(c(mNIDL,mNFLO),collapse=" and ")
        if(nchar(msg.extra)) {
            msg <- paste0(msg," (",msg.extra,").")
        } else {
            msg <- paste0(msg,".")
        }
        
        message(msg)
    }
    
    names(tables) <- meta[,name]


    tables <- lapply(tables,as.fun)
    
    writeNMinfo(tables,list(tables=meta),byRef=TRUE)
    
    return(tables)

}