berdox / apriori-algorithm-implementation

CPSC-4310-project

Usage

make all - Compiles Apriori, Idea1 and GenDatabase programs

Their respective targets are:

make Apriori
make Idea1
make GenDatabase

Cleaning the executables:

make clean

---

To create database files:

./GenDatabase

To run Apriori/Idea1 implementation:

./executable <filename> <min_sup>

example:

./Apriori ./databases/D1K.txt 0.01

Results

Apriori

D1K 0.15 - 2 DB Scans, 0.06s
D1K 0.10 - 3 DB Scans, 0.79s
D1K 0.08 - 3 DB Scans, 3.35s
D1K 0.05 - 3 DB Scans, 3.36s
D1K 0.01 - 4 DB Scans, 5.27s

D10K 0.15 - 2 DB Scans, 0.55s
D10K 0.10 - 3 DB Scans, 7.00s
D10K 0.08 - 3 DB Scans, 33.46s
D10K 0.05 - 3 DB Scans, 33.22s
D10K 0.01 - 4 DB Scans, 51.10s

D50K 0.15 - 2 DB Scans, 3.12s
D50K 0.10 - 3 DB Scans, 36.11s
D50K 0.08 - 3 DB Scans, 192.80s
D50K 0.05 - 3 DB Scans, 180.82s
D50K 0.01 - 4 DB Scans, 277.02s

D100K 0.15 - 2 DB Scans, 6.37s
D100K 0.10 - 3 DB Scans, 83.27s
D100K 0.08 - 3 DB Scans, 379.13s
D100K 0.05 - 3 DB Scans, 353.44s
D100K 0.01 - 4 DB Scans, 600.88s

Idea 1

D1K 0.15 - 1.0 DB Scans, 0.06s
D1K 0.10 - 2.0 DB Scans, 0.77s
D1K 0.08 - 2.0 DB Scans, 3.34s
D1K 0.05 - 1.7 DB Scans, 3.28s
D1K 0.01 - 1.1 DB Scans, 4.20s

D10K 0.15 - 1.0 DB Scans, 0.60s
D10K 0.10 - 2.0 DB Scans, 6.89s
D10K 0.08 - 1.9 DB Scans, 33.07s
D10K 0.05 - 1.5 DB Scans, 32.72s
D10K 0.01 - 1.1 DB Scans, 45.34s

D50K 0.15 - 1.0 DB Scans, 3.12s
D50K 0.10 - 2.0 DB Scans, 33.32s
D50K 0.08 - 1.8 DB Scans, 166.98s
D50K 0.05 - 1.5 DB Scans, 164.17s
D50K 0.01 - 1.1 DB Scans, 273.16s

D100K 0.15 - 1.0 DB Scans, 6.25s
D100K 0.10 - 2.0 DB Scans, 81.15s
D100K 0.08 - 1.8 DB Scans, 347.98s
D100K 0.05 - 1.5 DB Scans, 329.25s
D100K 0.01 - 1.1 DB Scans, 533.55s

apriori-algorithm-implementation's Issues

Implement Idea 1 as extension of original Apriori

When we go to implement idea 1, lets abstract away most of the Apriori class, and create an Idea1 class to extend it with the main method overridden.

Idea1 D100K at minSup .01 doesnt finish?

Running D100K at minSup 0.01 took me over 45 mins of runtime and didnt complete the second scan... maybe some complexity is blowing up here or the is an inf loop idx.