e4750_2022Fall_Project_LZSS_qz2487_bh2803

This is the repo for our final project of EECSE4750 Heterogeneous Computing. Our topic is LZSS parallel compression. We use a parallel implementation of LZSS algorithm to comrpess data.

File Organization

├── project.py
├── README.md
├── LZSSCPU.py
├── testfile
│   ├── big.txt
│   ├── random1mb.txt
│   ├── random5mb.txt
|   ├── random10mb.txt
│   └── random50mb.txt
├── result
│   └── ... (result bin files)
├── image
│   └── ... (result jpg files)
├── canterbury
│   ├── SHA1SUM
│   ├── alice29.txt
│   ├── cp.html
│   ├── fields.c
│   ├── grammar.lsp
│   ├── kennedy.xls
│   ├── lcet10.txt
│   ├── plrabn12.txt
│   ├── ptt5
│   ├── sum
│   └── xargs.1
├── kernel_old.py
└── kernel.py

Code running instruction:

1. kernel_old.py and kernel.py contains all the cuda kernel code in our project.
2. To test our code,

• run python project.py to test random number file with different sizes
• run python project.py -d to test different kinds of files
• add -a like python project.py -a to give a test on random number files with extra naive CPU test. Note: Naive implementation is very slow since it used dynamic arrays.
• add -f like python project.py -f FILENAME to test your own file. Note: DIY file doesn't give a plot back to folder image

File Details

1. Canterbury Corpus data is contained in canterbury folder. The Canterbury Corpus is a benchmark to enable researchers to evaluate lossless compression methods. This site includes test files and compression test results for many research compression methods.
2. testfile folder contains the random number data with differnet sizes.
3. result folder contains the result bin files which are not able to be opened. It can be decoded back to original file with proper implementation.
4. image folder returns the image file we produced during execution.

Plots with our test