Table of Contents
Introduction
PrefGen is a dataset generator with conditional preferences for StreamPref DSMS prototype. PrefGen generates the relations, queries and auxiliary files for the execution of experiments with StreamPref. The experiments parameters must be updated directly in the source code. The parameters are the following:
- ATT: Number of attributes;
- ATTRIBUTE_LIST: List with variation on parameter ATT;
- ATTRIBUTE_DEFAULT: Default value for parameter ATT;
- TUP: Number of initial tuples;
- TUPLE_LIST: List with variation on parameter ATT;
- TUPLE_DEFAULT: Default value for parameter ATT;
- DEL: Number of deletions per instant;
- DELETION_LIST: List with variation on parameter ATT;
- DELETION_DEFAULT: Default value for parameter ATT;
- INS: Number of insertions per instant;
- INSERTION_LIST: List with variation on parameter ATT;
- INSERTION_DEFAULT: Default value for parameter ATT;
- RUL: Number of rules;
- RULE_LIST: List with variation on parameter ATT;
- RULE_DEFAULT: Default value for parameter ATT;
- LEV: Maximum preference level;
- LEVEL_LIST: List with variation on parameter ATT;
- LEVEL_DEFAULT: Default value for parameter ATT;
- IND: Number of indifferent attributes;
- INDIFF_LIST: List with variation on parameter ATT;
- INDIFF_DEFAULT: Default value for parameter ATT;
- TOP: Number of top-k tuples to be returned.
- TOPK_LIST: List with variation on parameter ATT;
- TOPK_DEFAULT: Default value for parameter ATT;
The relations are composed of integer attributes. At every instant, INS tuples are inserted and DEL tuples are deleted. The number of iterations are controlled by the variable ITERATION_DEFAULT. Please see the related publications for more information.
Command Line
prefgen.py [-h] [-g] [-r] [-s]:
-h/--help: display help message
-g/--gen: Generate files
-r/--run: Run experiments
-s/--summarize: Summarize experiments results