Class MapRedScore

src/com/winvector/logistic/mr/MapRedScore.java:32–145

import com.winvector.variables.VariableEncodings;

public final class MapRedScore {
	private static final String MRFIELDNAME = "MapRedSc.MRBlock";
	public static final String IDEALHEADERFIELD = "MapRedSc.IdealHeader";
	private final SigmoidLossMultinomial underlying;
	private final WritableVariableList defs;
	private final boolean useIntercept;
	private final Configuration mrConfig;
	private final Path pathIn;

	
	public MapRedScore(final SigmoidLossMultinomial underlying, final WritableVariableList defs, final boolean useIntercept, 
			final Configuration mrConfig, final Path pathIn) {
		this.underlying = underlying;
		this.useIntercept = useIntercept;
		this.defs = defs;
		this.pathIn = pathIn;
		this.mrConfig = mrConfig;
	}
	
	
	public static final class ScoreMapper extends Mapper<LongWritable,Text,Text,Text> {
		private final NumberFormat nf = new DecimalFormat("000000000000");
		private LineBurster burster = null;
		// config
		private JobStateDescr config = null;
		// derived
		private VariableEncodings defs = null;
		private Log log = null;
		private String hostDescr = null;

		
		@Override
		public void setup(final Context context) throws IOException {
			log = LogFactory.getLog(ScoreMapper.class);
			hostDescr = WritableUtils.hostDescr();			
			log.info(".setup() " + hostDescr); 
			// read side-channel configuration
			try {
				burster = SerialUtils.readSerialiazlabeFromString(context.getConfiguration().get(MapRedScan.BURSTERSERFIELD));
			} catch (ClassNotFoundException e) {
				throw new IOException(e.toString());
			}
			config = JobStateDescr.fromString(context.getConfiguration().get(MRFIELDNAME));
			defs = new VariableEncodings(config.defs,config.useIntercept,config.weightKey);
		}

		@Override
		public void map(final LongWritable key, final Text value, final Context context) throws IOException, InterruptedException {
			final String origStr = value.toString();
			final BurstMap parsed = burster.parse(origStr);
			if(!parsed.isEmpty()) {
				final SparseSemiVec v = defs.vector(parsed);
				if(v!=null) {
					final double wt = defs.weight(parsed);
					final int catInt = -1;
					final Datum r = new SparseExampleRow(v,wt,catInt);
					final double[] pred = config.underlying.predict(config.x,r);
					final int argMax = HelperFns.argmax(pred);