plot-overlaps.cpp

Go to the documentation of this file.

#include "atomprobe/atomprobe.h"

using namespace std;
using namespace AtomProbe;

#include <fstream>
#include <set>
#include <iomanip>
#include <algorithm>

//This code demonstrates creating a "DOT" [1] graph, consisting of a plot of all the
// overlaps present in a rangefile.
//
// The graph's edges and its output show which species and masses overlap
// Note that there are lots and lots of output edges at the moment,
// so you may need to use the strict keyword in the output dot file (text file)
// These plots were initially developed by London, 2016.
//
// [1] https://en.wikipedia.org/wiki/DOT_language

int main(int argc, char *argv[])
{

    //Check input arguments
    if(argc < 2)
    {
        cerr << "USAGE : " << argv[0] << " (Rangefile| Ion list)" << endl;
        return 1;
    }

    RangeFile rng;
    if(argc == 2)
    {
        //Load the rangefile
        //---
        if(!rng.open(argv[1]))
        {
            cerr << "Failed to open rangefile :" << argv[1] << rng.getErrString() << endl;
            return 2;
        }
        cerr << "Rangefile has :" << rng.getNumIons() << " ranges, and " << rng.getNumIons() << " ions." << endl;
        //---
    }
    else
    {
        set<string> ionNames;
        //Fake a rangefile using the specified ions
        RGBf rgbf;
        for(unsigned int ui=1;ui<argc;ui++)
            ionNames.insert(string(argv[ui]));

        for(const auto & ionName : ionNames)
        {
            string s;
            s = ionName;

            rng.addIon(s.c_str(),s.c_str(),rgbf);
        }
    }

    //Open the table of masses of elements
    //---
    AbundanceData natTable;
    if(natTable.open("naturalAbundance.xml"))
    {
        cerr << "failed to open natural abundance file" << endl;
        return 3;
    }
    //---


    //Find the overlaps
    //---
    //Separation between overlap items
    const float MASS_DELTA=0.05;
    //Maximum charge to consider when evaluating overlaps
    const unsigned int MAX_CHARGE=3;
    vector<pair<size_t,size_t> > overlapIdx;
    vector<pair<float,float> > overlapMass;
    findOverlaps(natTable,rng,MASS_DELTA,MAX_CHARGE,overlapIdx,overlapMass);

    cerr << "Found :" << overlapIdx.size() << " overlaps, using max charge=" << MAX_CHARGE << " and mass delta=" << MASS_DELTA << endl;
    //---

    //Save the result into a DOT formatted file
    //---
    ofstream f("dotFile.txt");
    if(!f)
    {
        cerr << "Unable to open output file." <<endl;
        return 4;
    }

    // Get a list of all the nodes in the graph
    vector<pair<string, unsigned int> > nodeList;
    vector<string> nodeNames;
    for(unsigned int ui=0; ui<overlapIdx.size(); ui++)
    {
        // first entry of overlapIdx
        unsigned int rawId,id,charge;
        rawId =overlapIdx[ui].first;
        id = rawId/MAX_CHARGE;
        charge=rawId%MAX_CHARGE +1;
        nodeList.push_back( make_pair(rng.getName(id), charge) );

        // make a node label string for it
        // TODO: Add <SUB></SUB> around numbers for subscripts
        // TODO: replace only "1+" with "+", or replace "2+" with "++" etc.
        std::stringstream buffer;
        buffer << "\"" << rng.getName(id) << "|" << charge << "+\" ";
        buffer << " [ label = <" << rng.getName(id) << "<SUP>" << charge << "+</SUP>> ];" << endl;
        nodeNames.push_back(buffer.str());

        //second entry of overlapIdx
        rawId =overlapIdx[ui].second;
        id = rawId/MAX_CHARGE;
        charge=rawId%MAX_CHARGE+1;
        nodeList.push_back( make_pair(rng.getName(id), charge) );

        // make a node label string for 2nd node
        buffer.str(""); // clear buffer
        buffer << "\"" << rng.getName(id) << "|" << charge << "+\" ";
        buffer << " [ label = <" << rng.getName(id) << "<SUP>" << charge << "+</SUP>> ];" << endl;
        nodeNames.push_back(buffer.str());
    }

    // TODO: When two nodes share multiple edges (peaks), display them
    // as one edge with "peak1 \n peak2 \n peak3 etc." rather than
    // multiple edges.

    // start graph
    f << "strict graph overlap_map {" << endl;

    // Only print unique labels to cut down file size
    // Find unique strings using a set
    set <std::string> uniqueStrings;
    for (auto & nodeName : nodeNames)
    {
        uniqueStrings.insert(nodeName);
    }
    // Print the unique labels
    for (const auto & uniqueString : uniqueStrings)
    {
        f << uniqueString;
    }

    for(unsigned int ui=0; ui<overlapIdx.size(); ui++)
    {
        //Create graph edge
        f << "\"" << nodeList[ui*2].first << "|" << nodeList[ui*2].second << "+\" -- ";

        f << "\"" << nodeList[ui*2+1].first << "|" << nodeList[ui*2+1].second << "+\"";

        // average mass to label the edge with
        float avgMass = (overlapMass[ui].first + overlapMass[ui].second)/2.0;
        f << " [ label = \"" << std::fixed << std::setprecision( 1 ) << avgMass << "\" ];" << endl;
    }
    f << "}" << endl;
    //---

    cerr << "Wrote overlaps to dotFile.txt" << endl;
    return 0;
}