mobios.app
Class BuildVPIndex

java.lang.Object
  mobios.app.BuildVPIndex

public class BuildVPIndex
extends Object

This is the main entry point for building a VP Index. It can build either a single VPIndex, or a series of VPIndexes.

Version:

2006.07.26

Author:

Rui Mao, Willard

Method Summary

static void

main(String[] args) The basic commandline options are: -t [data type: one of "protein", "dna", "vector", "image", "ms", "msms"] -d [location of "mobiosData" directory; if directory does not exist, will be created at specified location] -i [input data file name] -o [output index file name] -psm [the pivot selection selection method: random, fft, center, pcaonfft, pca] -p [number of pivots in an index node] -dpm [data partition method: balanced, clusteringkmeans, clusteringboundary] -f [fanout of a pivot] -m [maximum number of children in leaf nodes] -pl [path length, default 0] -g [debug level] -frag [fragment length, only meaningful for sequences] -dim [dimension of vector data to load] -b [whether bucketing will be used, 1: use] -s [size of index] -r [maximum radius for partition] For building multiple databases, use the following options: -sm [size of smallest index] -la [size of largest index] -st [step size of index] When using the advanced options, the size of the index will be appended to the given index name for each index.

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Method Detail

main

public static void main(String[] args)

The basic commandline options are: -t [data type: one of "protein", "dna", "vector", "image", "ms", "msms"] -d [location of "mobiosData" directory; if directory does not exist, will be created at specified location] -i [input data file name] -o [output index file name] -psm [the pivot selection selection method: random, fft, center, pcaonfft, pca] -p [number of pivots in an index node] -dpm [data partition method: balanced, clusteringkmeans, clusteringboundary] -f [fanout of a pivot] -m [maximum number of children in leaf nodes] -pl [path length, default 0] -g [debug level] -frag [fragment length, only meaningful for sequences] -dim [dimension of vector data to load] -b [whether bucketing will be used, 1: use] -s [size of index] -r [maximum radius for partition] For building multiple databases, use the following options: -sm [size of smallest index] -la [size of largest index] -st [step size of index] When using the advanced options, the size of the index will be appended to the given index name for each index. The Metric is hardcoded for each data type.

Parameters:

args -