mrjoy/hadoop-qctree

Quotient Cube Implementation using Hadoop

hadoop-qctree

QC-tree(short for quotient cube tree) is a compact data structure for representing and implementing quotient cube.

QC-tree

• retain all the essential information in a quotient lattice, yet be concise
• enable efficient answering of various kinds of queries including point, range,and iceberg queries
• afford efficient maintenance against updates

hadoop-qctree provides QC-tree construction using Hadoop.

How to Build Binary

To build the binary of the hadoop-qctree, run the following command

mvn clean package

The above command produces a tar ball in the target directory.

How to Run

Local Mode

Untar the tar ball and move to hadoop-qctree-* directory. Need to update the table.json in the conf directory as per input structure

To build a qc-tree for the given input, execute the following command

bin/qctree.sh input-file output-dir

Once QC-tree construction is over, to run any query execute the following command

bin/query.sh output-of-previous-job output-dir query-tobe-run

As of now, query should be given as a single string, dimensions separated by comma(,)

Example query: "S2,*,s"