INSTALL

INSTALLING PARALLPAIRS
================

Install dependencies first.

The package uses the following ocaml packages: ocamlgraph unix str threads bigarray gz mpi

Some of these may not be used in parallpairs which is a subset of a larger text processing package.

Then, simply run:

$ cd src
$ make

You can test the compilation with:

$ src/all-pairs -help
Usage: all-pairs -dv <dv-file> -threshold <sim-threshold> -algo <algorithm-number> [-list]
  -dv document vector
  -threshold similarity threshold
  -sigma frequent term cutoff threshold
  -th epsilon_h
  -algo algorithm to run
  -nonorm do not normalize
  -list list algorithms
  -help  Display this list of options
  --help  Display this list of options
Rigel:parallpairs malfunct$ src/all-pairs -list
0: all-pairs-0
1: all-pairs-0-array
2: all-pairs-0-array2
3: all-pairs-0-remscore
4: all-pairs-0-minsize
5: all-pairs-1
6: all-pairs-1-dense
7: all-pairs-1-remscore
8: all-pairs-1-upperbound
9: all-pairs-1-minsize
10: all-pairs-1_remscore_minsize
11: all-pairs-2
12: all-pairs-bruteforce
13: all-pairs-cutoff
14: all-pairs-cutoff2

$ src/all-pairs -dv ../data-github/dv/radikal.dv -threshold 0.9 -algo 1
Reading DV file ../data-github/dv/radikal.dv
Normalizing

* Running all-pairs-0-array
doc 6800     
Time elapsed: 8.505927
Number of matches: 10

(The dataset is on github)