Add README.org

* README.org: New file.

DESCRIPTION

@@ -20,4 +20,6 @@ Collate:
     'mix.R'
     'run.R'
 Imports: 
-    RMassBank
+    RMassBank,
+    parallel,
+    yaml

R/mix.R

@@ -4,6 +4,7 @@ stripext<-function(fn) {
 
 ##' Create directories without drama.
 ##'
+##' Create directories without drama.
 ##' 
 ##' @title Create directories without drama
 ##' @param path Name of the directory.
@@ -26,19 +27,39 @@ no_drama_mkdir<-function(path) {
 ##' @param file The name of the YAML specification that will be merged
 ##'     with the template Rmb settings file.
 ##' @return NULL
-##' @author Todor Kondić
 mk_sett_file<-function(sett_alist,file) {
-    require(yaml)
     tmp<-tempfile()
     RMassBank::RmbSettingsTemplate(tmp)
-    sett<-yaml.load_file(tmp)
+    sett<-yaml::yaml.load_file(tmp)
     for (nm in names(sett_alist)) {
         sett[[nm]]<-sett_alist[[nm]]
     }
-    write_yaml(x=sett,file=file)
+    yaml::write_yaml(x=sett,file=file)
     NULL
 }
 
+##' Combine the RMB settings files
+##' 
+##' Combine RMB settings with different collisional energies into one
+##' settings file with multiple collisional energy entries.
+##' 
+##' @title Combine RMB Settings With Different Collisional Energies
+##' @param sett_fns A list of settings files.
+##' @param fname The name of the combined file.
+##' @return fname
+##' @author Todor Kondić
+mk_combine_file<-function(sett_fns,fname) {
+    all_settings <- lapply(sett_fns,yaml::yaml.load_file)
+    comb_settings <- all_settings[[1]]
+    
+    for (n in 1:length(all_settings)) {
+        comb_settings$spectraList[[n]] <- all_settings[[n]]$spectraList[[1]]
+    }
+
+    yaml::write_yaml(x=comb_settings,fname)
+    fname
+}
+
 ##' Generate the RMassBank compound list from the input compound list
 ##' in CSV file src_fn. The input compound list format is either a
 ##' Chemical Dashboard csv file with, at least, PREFERRED_ SMILES
@@ -86,14 +107,12 @@ gen_comp_list<-function(src_fn,dest_fn) {
 ##'     documentation of `msmsRead` for details.
 ##' @param archdir The directory to store R objects created during
 ##'     workflow execution.
+##' @param lastStep The last step in the workflow. Default is eight.
 ##' @return MsmsWorkspace object.
 ##' @author Todor Kondić
-single.sw<-function(fn_data,stgs_alist,wd,fn_cmpd_list,mode,readMethod="mzR",archdir="archive") {
-    
-    require(RMassBank)
-    require(yaml)
+single.sw<-function(fn_data,stgs_alist,wd,fn_cmpd_list,mode,readMethod="mzR",archdir="archive",lastStep=8) {
     ## Generate settings file and load.
-    stgs_alist<-if (is.character(stgs_alist)) yaml.load_file(stgs_alist) else stgs_alist
+    stgs_alist<-if (is.character(stgs_alist)) yaml::yaml.load_file(stgs_alist) else stgs_alist
     sfn<-file.path(wd,paste(fn_data,".ini",sep=''))
     mk_sett_file(stgs_alist,sfn)
     RMassBank::loadRmbSettings(sfn)
@@ -109,14 +128,14 @@ single.sw<-function(fn_data,stgs_alist,wd,fn_cmpd_list,mode,readMethod="mzR",arc
     write.csv(x=df_table,file=fn_table,row.names=F)
 
     ## Make empty workspace.
-    w <- newMsmsWorkspace()
+    w <- RMassBank::newMsmsWorkspace()
     ## Run the workflow.
     message(paste("Reading in file:",fn_data))
-    w <-msmsRead(w,filetable=fn_table,readMethod="mzR",mode=mode)
+    w <-RMassBank::msmsRead(w,filetable=fn_table,readMethod="mzR",mode=mode)
     archdir<-file.path(wd,archdir)
     if (!dir.exists(archdir)) dir.create(archdir)
     fn_arch<-file.path(archdir,paste(fn_data,".archive",sep=''))
-    RMassBank::msmsWorkflow(w, mode=mode, steps=2:8,archivename=fn_arch)
+    RMassBank::msmsWorkflow(w, mode=mode, steps=2:lastStep,archivename=fn_arch)
 }
 
 
@@ -193,14 +212,95 @@ mb.single<-function(mb,infodir,fn_stgs) {
 ##' @param mode Same as in msmsRead.
 ##' @param readMethod Same as in msmsRead.
 ##' @param archdir Name of the archive.
+##' @param lastStep The last step of the spectral workflow.
+##' @param combine If TRUE, use combineMultiplicies to merge
+##'     workspaces corresponding to different collisional energies.
 ##' @return A named list of spectral workspaces. The names are derived
 ##'     from data filenames.
 ##' @author Todor Kondić
-v<-function(fn_data,stgs_alist,wd,fn_cmpd_list,mode,readMethod="mzR",archdir="archive") {
+v<-function(fn_data,stgs_alist,wd,fn_cmpd_list,mode,readMethod="mzR",archdir="archive",lastStep=8,combine=F) {
+    idir<-function(n) file.path(".",stripext(n))
     f<-Vectorize(single.sw,vectorize.args=c("wd","fn_data","stgs_alist"),SIMPLIFY=F)
-    x<-f(fn_data,stgs_alist,wd,fn_cmpd_list,mode,readMethod=readMethod,archdir=archdir)
-    names(x)<-basename(fn_data)
-    x}
+    rootdir <- getwd()
+    if (combine) {
+        z<-f(fn_data,stgs_alist,wd,fn_cmpd_list,mode,readMethod=readMethod,archdir=archdir,lastStep=7)
+        names(z)<-basename(fn_data)
+        zz<-RMassBank::combineMultiplicities(z)
+
+        combdir<-"combined"
+        archdir<-file.path(rootdir,combdir,archdir)
+        no_drama_mkdir(combdir)
+        no_drama_mkdir(archdir)
+        fn_arch<-file.path(archdir,"archive")
+        fn_comb_stgs <- file.path(rootdir,combdir,paste(combdir,".mzML.ini",sep=''))
+        ddirs <- sapply(names(z),idir)
+        stgs_fls <- sapply(ddirs,function(x) file.path(x,paste(x,".mzML.ini",sep='')))
+        mk_combine_file(stgs_fls,fn_comb_stgs)
+
+        res<-list(RMassBank::msmsWorkflow(zz, steps=8, mode=mode, archivename = fn_arch))
+        names(res)<-paste(combdir,".mzML",sep='') #Clearly a hack.
+        res
+    } else {
+        z<-f(fn_data,stgs_alist,wd,fn_cmpd_list,mode,readMethod=readMethod,archdir=archdir,lastStep=lastStep)
+        names(z)<-basename(fn_data)
+        z
+    }
+}
+
+##' Interface to parallel spectral workflow.
+##'
+##' 
+##' @title Parallel Spectral Workflow.
+##' @param fn_data A sequence of mzML input files.
+##' @param stgs_alist A list of named list of settings, or a list of
+##'     filenames of YAML files containing the settings.
+##' @param wd The list of working directories.
+##' @param fn_cmpd_list The compound list characterising the mixtures.
+##' @param mode Same as in msmsRead.
+##' @param readMethod Same as in msmsRead.
+##' @param archdir Name of the archive.
+##' @param lastStep The last step in spectral workflow.
+##' @param combine If TRUE, use combineMultiplicies to merge
+##'     workspaces corresponding to different collisional energies.
+##' @param cl Cluster.
+##' @return A named list of spectral workspaces. The names are derived
+##'     from data filenames.
+##' @author Todor Kondić
+p.sw<-function(fn_data,stgs_alist,wd,fn_cmpd_list,mode,readMethod="mzR",archdir="archive",lastStep=8,combine=F,cl=NULL) {
+    idir<-function(n) file.path(".",stripext(n))
+    fnocomb<-function(fn,stgs,wd) {
+        single.sw(fn,stgs,wd,fn_cmpd_list,mode,readMethod,archdir,lastStep=lastStep)
+    }
+    fcomb<-function(fn,stgs,wd) {
+        single.sw(fn,stgs,wd,fn_cmpd_list,mode,readMethod,archdir,lastStep=7)
+    }
+
+    if (combine) {
+        rootdir <- getwd()
+        z<-parallel::clusterMap(cl,fcomb,fn_data,stgs_alist,wd)
+        names(z)<-basename(fn_data)
+        zz<-RMassBank::combineMultiplicities(z)
+
+        combdir<-"combined"
+        archdir<-file.path(rootdir,combdir,archdir)
+        no_drama_mkdir(combdir)
+        no_drama_mkdir(archdir)
+        fn_arch<-file.path(archdir,"archive")
+        fn_comb_stgs <- file.path(rootdir,combdir,paste(combdir,".mzML.ini",sep=''))
+        ddirs <- sapply(names(z),idir)
+        stgs_fls <- sapply(ddirs,function(x) file.path(x,paste(x,".mzML.ini",sep='')))
+        mk_combine_file(stgs_fls,fn_comb_stgs)
+        
+        res<-list(RMassBank::msmsWorkflow(zz, steps=8, mode=mode, archivename = fn_arch))
+        names(res)<-paste(combdir,".yml",sep='') #Clearly a hack.
+        res
+    } else {
+        z<-parallel::clusterMap(cl,fnocomb,fn_data,stgs_alist,wd)
+        names(z)<-basename(fn_data)
+        z
+    }
+}
+
     
 ##' Interface to vectorised Mass Bank workflow.
 ##'
@@ -217,3 +317,20 @@ mb.v<-function(mb,infodir,fn_stgs) {
     x<-f(mb,infodir,fn_stgs)
     names(x)<-names(mb)
     x}
+
+##' Interface to parallelised Mass Bank workflow.
+##'
+##' 
+##' @title Parallel Mass Bank Workflow
+##' @param mb List of mass bank workflow objects
+##' @param infodir List of subdirs containing info lists.
+##' @param fn_stgs List of settings files.
+##' @param cl Cluster.
+##' @return A named list of mbWorkspace objects. The names are derived
+##'     from the input mb sequence.
+##' @author Todor Kondić
+mb.p<-function(mb,infodir,fn_stgs,cl=F) {
+    x<-parallel::clusterMap(cl=cl,mb.single,mb,infodir,fn_stgs)    
+    names(x)<-names(mb)
+    x}
+

R/run.R

@@ -29,16 +29,25 @@ attch<-function(...) paste(...,sep='')
 ##' @param fn_cmpd_list Compound list.
 ##' @param mode as in msmsRead.
 ##' @param rdir The root data directory.
+##' @param combine If TRUE, use combineMultiplicies to merge
+##'     workspaces corresponding to different collisional energies.
+##' @param proc Split work between this amount of processes. If FALSE
+##'     (or, 1), run sequential.
 ##' @return A named list of msmsWorkspace objects.
 ##' @author Todor Kondić
 ##' @export
-sw.do<-function(fn_data,fn_cmpd_list,mode,rdir=".") {
- 
+sw.do<-function(fn_data,fn_cmpd_list,mode,rdir=".",combine=F,proc=F) {
     no_drama_mkdir(rdir)
     wdirs<-sapply(basename(fn_data),function(nm) file.path(rdir,stripext(nm)))
     sapply(wdirs,no_drama_mkdir)
     stgs<-sapply(basename(wdirs),function (nm) paste(nm,"yml",sep='.'))
-    v(fn_data,stgs,wdirs,fn_cmpd_list,mode)
+
+    if (proc) {
+        cl<-parallel::makeCluster(proc)
+        p.sw(fn_data,stgs,wdirs,fn_cmpd_list,mode,combine=combine,cl=cl)
+    } else {
+        v(fn_data,stgs,wdirs,fn_cmpd_list,mode,combine=combine)
+    }
 }
 
 ##' Creates and prepares mbWorkspace objects before the full workflow
@@ -50,6 +59,7 @@ sw.do<-function(fn_data,fn_cmpd_list,mode,rdir=".") {
 ##' @title Prepare mbWorkspace objects
 ##' @param w A list of spectral workspace inputs.
 ##' @param rdir Data root.
+##' @param proc Split work between this amount of processes. If FALSE
 ##' @return Named list of prepared mbWorkspace objects.
 ##' @author Todor Kondić
 ##' @export
@@ -70,12 +80,20 @@ mb.prep<-function(w,rdir=".") {
 ##' @title Perform the Mass Bank workflow
 ##' @param mb The list of prepared mbWorkspace objects.
 ##' @param rdir Root data dir.
+##' @param proc Split work between this amount of processes. If FALSE
+##'     (or, 1), run sequential.
 ##' @return The named list of processed mbWorkspace objects.
 ##' @author Todor Kondić
 ##' @export
-mb.do<-function(mb,rdir=".") {
+mb.do<-function(mb,rdir=".",proc=F) {
     idir<-function(n) file.path(rdir,stripext(n))
     infodir<-sapply(names(mb),function(n) file.path(idir(n),"info"))
     fn_stgs<-sapply(names(mb),function(n) file.path(idir(n),attch(n,'.ini')))
-    mb.v(mb,infodir,fn_stgs)
+
+    if (proc) {
+        cl<-parallel::makeCluster(proc)
+        mb.p(mb,infodir,fn <- stgs,cl=cl)
+    } else {
+        mb.v(mb,infodir,fn_stgs)
+    }
 }

man/mb.do.Rd

@@ -4,12 +4,15 @@
 \alias{mb.do}
 \title{Perform the Mass Bank workflow}
 \usage{
-mb.do(mb, rdir = ".")
+mb.do(mb, rdir = ".", proc = F)
 }
 \arguments{
 \item{mb}{The list of prepared mbWorkspace objects.}
 
 \item{rdir}{Root data dir.}
+
+\item{proc}{Split work between this amount of processes. If FALSE
+(or, 1), run sequential.}
 }
 \value{
 The named list of processed mbWorkspace objects.

man/mb.p.Rd

+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/mix.R
+\name{mb.p}
+\alias{mb.p}
+\title{Parallel Mass Bank Workflow}
+\usage{
+mb.p(mb, infodir, fn_stgs, cl = F)
+}
+\arguments{
+\item{mb}{List of mass bank workflow objects}
+
+\item{infodir}{List of subdirs containing info lists.}
+
+\item{fn_stgs}{List of settings files.}
+
+\item{cl}{Cluster.}
+}
+\value{
+A named list of mbWorkspace objects. The names are derived
+from the input mb sequence.
+}
+\description{
+Interface to parallelised Mass Bank workflow.
+}
+\author{
+Todor Kondić
+}

man/mb.prep.Rd

@@ -10,6 +10,8 @@ mb.prep(w, rdir = ".")
 \item{w}{A list of spectral workspace inputs.}
 
 \item{rdir}{Data root.}
+
+\item{proc}{Split work between this amount of processes. If FALSE}
 }
 \value{
 Named list of prepared mbWorkspace objects.

man/mk_combine_file.Rd

+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/mix.R
+\name{mk_combine_file}
+\alias{mk_combine_file}
+\title{Combine RMB Settings With Different Collisional Energies}
+\usage{
+mk_combine_file(sett_fns, fname)
+}
+\arguments{
+\item{sett_fns}{A list of settings files.}
+
+\item{fname}{The name of the combined file.}
+}
+\value{
+fname
+}
+\description{
+Combine the RMB settings files
+}
+\details{
+Combine RMB settings with different collisional energies into one
+settings file with multiple collisional energy entries.
+}
+\author{
+Todor Kondić
+}

man/mk_sett_file.Rd

@@ -20,6 +20,3 @@ Produce the Rmb Settings file
 Produce the Rmb Settings file based on the customisation file in
 YAML format.
 }
-\author{
-Todor Kondić
-}

man/no_drama_mkdir.Rd

@@ -15,6 +15,9 @@ The character string containing the input argument \code{path}.
 \description{
 Create directories without drama.
 }
+\details{
+Create directories without drama.
+}
 \author{
 Todor Kondić
 }

man/p.sw.Rd

+% Generated by roxygen2: do not edit by hand
+% Please edit documentation in R/mix.R
+\name{p.sw}
+\alias{p.sw}
+\title{Parallel Spectral Workflow.}
+\usage{
+p.sw(fn_data, stgs_alist, wd, fn_cmpd_list, mode, readMethod = "mzR",
+  archdir = "archive", lastStep = 8, combine = F, cl = NULL)
+}
+\arguments{
+\item{fn_data}{A sequence of mzML input files.}
+
+\item{stgs_alist}{A list of named list of settings, or a list of
+filenames of YAML files containing the settings.}
+
+\item{wd}{The list of working directories.}
+
+\item{fn_cmpd_list}{The compound list characterising the mixtures.}
+
+\item{mode}{Same as in msmsRead.}
+
+\item{readMethod}{Same as in msmsRead.}
+
+\item{archdir}{Name of the archive.}
+
+\item{lastStep}{The last step in spectral workflow.}
+
+\item{combine}{If TRUE, use combineMultiplicies to merge
+workspaces corresponding to different collisional energies.}
+
+\item{cl}{Cluster.}
+}
+\value{
+A named list of spectral workspaces. The names are derived
+from data filenames.
+}
+\description{
+Interface to parallel spectral workflow.
+}
+\author{
+Todor Kondić
+}

man/single.sw.Rd

@@ -5,7 +5,7 @@
 \title{RMassBank Spectral Workflow on a Single Compound Mixture}
 \usage{
 single.sw(fn_data, stgs_alist, wd, fn_cmpd_list, mode,
-  readMethod = "mzR", archdir = "archive")
+  readMethod = "mzR", archdir = "archive", lastStep = 8)
 }
 \arguments{
 \item{fn_data}{A mzML data file.}
@@ -26,6 +26,8 @@ documentation of \code{msmsRead} for details.}
 
 \item{archdir}{The directory to store R objects created during
 workflow execution.}
+
+\item{lastStep}{The last step in the workflow. Default is eight.}
 }
 \value{
 MsmsWorkspace object.

man/sw.do.Rd

@@ -4,7 +4,7 @@
 \alias{sw.do}
 \title{Perform MassBank Workflow on Multiple Compound Mixtures}
 \usage{
-sw.do(fn_data, fn_cmpd_list, mode, rdir = ".")
+sw.do(fn_data, fn_cmpd_list, mode, rdir = ".", combine = F, proc = F)
 }
 \arguments{
 \item{fn_data}{List of mzML data filenames to be processed.}
@@ -14,6 +14,12 @@ sw.do(fn_data, fn_cmpd_list, mode, rdir = ".")
 \item{mode}{as in msmsRead.}
 
 \item{rdir}{The root data directory.}
+
+\item{combine}{If TRUE, use combineMultiplicies to merge
+workspaces corresponding to different collisional energies.}
+
+\item{proc}{Split work between this amount of processes. If FALSE
+(or, 1), run sequential.}
 }
 \value{
 A named list of msmsWorkspace objects.

man/v.Rd

@@ -5,7 +5,7 @@
 \title{Vectorised Spectral Workflow.}
 \usage{
 v(fn_data, stgs_alist, wd, fn_cmpd_list, mode, readMethod = "mzR",
-  archdir = "archive")
+  archdir = "archive", lastStep = 8, combine = F)
 }
 \arguments{
 \item{fn_data}{A sequence of mzML input files.}
@@ -22,6 +22,11 @@ filenames of YAML files containing the settings.}
 \item{readMethod}{Same as in msmsRead.}
 
 \item{archdir}{Name of the archive.}
+
+\item{lastStep}{The last step of the spectral workflow.}
+
+\item{combine}{If TRUE, use combineMultiplicies to merge
+workspaces corresponding to different collisional energies.}
 }
 \value{
 A named list of spectral workspaces. The names are derived

mix.R

-## Create a temporary directory to hold the files generated on the
-## fly.
-rmbmix.mkdir<-function() {
-    nm<-tempfile(pattern="rmbmix")
-    dir.create(nm)
-    nm
-}
-
-##
-rmbmix.mk_sett_file<-function(sett_alist,file) {
-    require(yaml)
-    tmp<-tempfile()
-    RmbSettingsTemplate(tmp)
-    sett<-yaml.load_file(tmp)
-    for (nm in names(sett_alist)) {
-        sett[[nm]]<-sett_alist[[nm]]
-    }
-    write_yaml(x=sett,file=file)
-}
-
-
-## Generate the RMassBank compound list from the input compound list
-## in CSV file src_fn. The input compound list format is either
-## Chemical Dashboard csv file with, at least, PREFERRED_NAMES and
-## SMILES columns _filled_ out, or just an ordinary csv file with
-## columns SMILES and Names filled. Argument dest_fn is the
-## destination filename. Returns the number of compounds.
-rmbmix.gen_comp_list<-function(src_fn,dest_fn) {
-    df<-read.csv(src_fn)
-    ## Names
-    nms<-if ("PREFERRED_NAME" %in% names(df)) df$PREFERRED_NAME else df$Name
-    if (is.null(nms)) stop("Unable to read compound names from the input compound list.")
-
-    ## SMILES
-    haha<-df$SMILES
-    sz<-length(haha)
-    
-    ## CAS
-    casvals<-if ("CASRN" %in% names(df)) df$CASRN else rep(NA,sz)
-    if (is.null(haha)) stop("Unable to read SMILES from the input compound list.")
-
-    outdf<-data.frame(ID=1:sz,Name=nms,SMILES=haha,CAS=casvals,RT=rep(NA,sz))
-    write.csv(outdf,file=dest_fn,row.names=F,na="")
-    length(nms)
-}
-    
-    
-    
-## Perform the compound mixture workflow on the data file called
-## fn_data with settings named list called stgs_alist. Alternatively,
-## stg_alist can be a file name which follows the RMassBank settings
-## specification, also in YAML format, containing only parts that
-## differ from the default. Argument fn_cmpd_list is the compound
-## list. Argument wd is the scratch dir to hold generated ini files
-## and the like. Arguments mode and readMethod are the same as in
-## msmsRead.
-rmbmix.single<-function(fn_data,stgs_alist,fn_cmpd_list,wd,mode,readMethod="mzR",archdir="archive") {
-    
-    require(RMassBank)
-    require(yaml)
-    ## Generate settings file and load.
-    stgs_alist<-if (is.character(stgs_alist)) yaml.load_file(stgs_alist) else stgs_alist
-    sfn<-file.path(wd,paste(fn_data,".ini",sep=''))
-    rmbmix.mk_sett_file(stgs_alist,sfn)
-    loadRmbSettings(sfn)
-
-    ## Generate and load the compound list.
-    fn_comp<-file.path(wd,paste(fn_data,".comp.csv",sep=''))
-    n_cmpd<-rmbmix.gen_comp_list(fn_cmpd_list,fn_comp)
-    loadList(fn_comp)
-
-    ## Generate file table.
-    df_table<-data.frame(Files=rep(fn_data,n_cmpd),ID=1:n_cmpd)
-    fn_table<-file.path(wd,paste("fn-table.",fn_data,".csv",sep=''))
-    write.csv(x=df_table,file=fn_table,row.names=F)
-
-    ## Make empty workspace.
-    w <- newMsmsWorkspace()
-    ## Run the workflow.
-    message(paste("Reading in file:",fn_data))
-    w <-msmsRead(w,filetable=fn_table,readMethod="mzR",mode=mode)
-    if (!dir.exists(archdir)) dir.create(archdir)
-    fn_arch<-file.path(archdir,paste(fn_data,".archive",sep=''))
-    w<-msmsWorkflow(w, mode=mode, steps=2:8,archivename=fn_arch)
-    mb<-newMbWorkspace(w)
-    mb<-resetInfolists(mb)
-    ## loadInfolists
-    ## addPeaks
-    bits<-strsplit(fn_data,split="\\.")[[1]]
-    fn_info<-if (length(bits)> 1) paste(head(bits,-1),collapse=".") else fn_data
-    infodir<-fn_info
-    fn_info<-paste(archdir,"/",fn_info,".csv",sep='')
-    mb<-mbWorkflow(mb,infolist_path=fn_info)
-    list(w=w,mb=mb)
-}
-
-
-rmbmix.mb2.single<-function(mb,infodir) {
-    mb <- resetInfolists(mb)
-    mb <- loadInfolists(mb,infodir)
-    mbWorkflow(mb)
-}
-rmbmix.mb2<-Vectorize(rmbmix.mb2.single,vectorize.args=c("mb","infodir"),SIMPLIFY=F)
-rmbmix<- Vectorize(rmbmix.single,vectorize.args=c("fn_data","stgs_alist"),SIMPLIFY=F)