| Current Forum: Homework 4 - Huffman Trees (Part 1) |
| Date: Thu Oct 4 2001 10:20 pm |
| Author: Bortz, Andrew S. <abortz@andrew.cmu.edu> |
| Subject: Best way to store the code table |
|
|
| Ultimately, a key part of Huffman coding is how to reconstruct the tree/code table on the other side. The simplest if of course to store the frequency information, which allows the decompressor to create the Huffman tree the same way it was created during the compression. But the question is, which way is most efficient? The requirement of storing the code table/tree can kill a compression ratio if the file being compressed is very small, since the code table itself takes up a large percentage of the file size. So the size of header is what determines efficiency in this problem. |
|