CS 197 Hash Table project
The goal of this project is to produce an adaptive hash table library to make hash table implementations in read-write databases more effecient.
Note: The Makefile now works! You can still use the one-liner in MakefileLite if you wish.
Methods:
We are currently using hash tables implemented in C++. Since C++ has efficiency guarantees and was the language used by our nearest neighbor paper called A Seven Dimensional Analysis of Hashing Methods and Its Implications on the Querying Process which can be found here: http://www.vldb.org/pvldb/vol9/p96-richter.pdf, we decided it was the best choice. The source code for the hashtables, which are all of the form [Table Name Here].hpp all come from our parent paper's authors and was graciously sent to us by Professor Dittrich.
Our novel contribution is the construction of a hashtable implementation which can dynamically switch hashscheme/hashfunction combination depending on the given situation. Data Generation
Data generation for 2 data distrubutions are supported - dense and sparse.
dense is defined as keys [1,n] for a given n.
sparse is defined as n random keys in the range [1,n]
ed
data.py is the file to generate data sets. Usage is as follows:
python3 data.py -[sparse/dense] n.
This will write to a text file, either dense.txt or sparse.txt given the parameters.