pandas >= 0.24.1
ubuntu 18.01
python3 main.py -i path/to/file.csv -m minimun\_support -n c minimum\_confidence -n numrule
python3 benchmark.py
for benchmark, confidence is set to 0, and support is from 0.1 to 0.5 (where there are no rules with higher support).
python3 test/test\_script.py
- core: core package of the project, implementation of apriori algorithm and its supporting data structure
- core.database: abstract class for our database. provide function to count support, confident, etc.
- core.csv_database: implementation to use with csv file
- core.attr_value: fundamental item unit, defined over an column and its attribute.
- core.itemset: defined the itemset
- core.rule: define the rule
- core.apriori: implementation of apriori algorithm
- test: modules containing all unittests
- test_data: containing dataset used for testing and benchmark.
- main.py: main file for running
- benchmark.py: main file for running benchmark.
- running time is very high (running benchmark for details)
Support 0.10000, runtime 390.110582892, numrules 6090
Support 0.10404, runtime 370.372849130, numrules 5964
Support 0.10808, runtime 332.196668411, numrules 5700
...