I am trying to implement Golomb coding, but I don't understand how it's tuned to obtain optimal code.

It is said that

Golomb coding uses a tunable parameter M to divide an input value into two parts: q, the result of a division by M, and r, the remainder. The quotient is sent in unary coding, followed by the remainder in truncated binary encoding.

I don't understand how should I choose the parameter M - I can't see how the explanation in Wikipedia relates to actual data. I believe it should be related to statistical moments, is that true?

For example, if I have this example set:
I believe M should be very small for this kind of data. I bet it's either 1 or 2. It's mean is ~2.2 and standard deviation is ~1.1. My intuition would tell me to choose 2.

Another dataset here:
This time the mean is ~7.2 and standard deviation is ~5.0.

Is 7 the right value in this case? And should I prefer Rice code (use 8 as it is a power of 2) if I get a value like 7?

I understand that division will be easier if I use Rice coding, but are there any benefits in NOT using it? I mean - 3 bits will be used for remainder in either case, how could pure Golomb code be more optimal then?

One more nuance - Golomb code is for nonnegative integers. If I have positive integers instead, should I save x-1 instead? It would change a lot for the first of the mentioned datasets.

  • 1
    Have you tried asking this question on another SO site? I'm not saying it definitely doesn't belong here, just that you might find more expertise for an answer somewhere else. – Encaitar May 14 '14 at 7:49
  • @Encaitar, seems I'll have to reask... Which site do you think would be better for this? Mathematics or stackoverflow? – Džuris May 16 '14 at 17:51
  • I'm not sure if there is a rule against posting to both. StackOverflow would properly get more people looking at it, and their moderators will recommend moving it to another site if they think that is best. – Encaitar May 18 '14 at 22:00

The Golomb-Rice algorithm doesn't specify how to find the optimal parameters, and in the general case you will have to try to infer the posterior probability of symbol occurrences in the dataset to estimate the optimal value of M. Note that it is a common choice to have M = 2 k, a power of 2, as coding in this case is simple, and then discuss k instead. The search for the optimal k is usually done exhaustively over the dataset.

Following the above you can now understand why the wikipedia article you link to doesn't refer to any actual data, but says the follows as a way of an example,

Given an alphabet of two symbols, or a set of two events, P and Q, with probabilities p and (1 − p) respectively, where p ≥ 1/2, Golomb coding can be used to encode runs of zero or more P's separated by single Q's. In this application, the best setting of the parameter M is the nearest integer to \frac{-1}{\log_{2}p}.

In the case where the probability p is known, where we have the prior equivalent to the posterior distribution, it can be shown that there is a best known value for M, but not otherwise.

Practically what this means is that you have to have a fairly good idea of how your eventual dataset will look like, possibly further increasing confidence with re-sampling methods (e.g., bootstrapping), and then searching exhaustively1 for the optimal parameter value in the sample datasets - the k that minimises the expected code length. Then you use an average value of the k you've decided on for the future datasets. Some implementations store a table of input dataset characteristics and adaptively select (i.e. change) the code parameters when they detect that the input pattern shifts. For example it is common to evaluating the running characteristics of the input sequence, mean, variance, etc, and select accordingly when thresholds are breached.

1 There are precise bounds that could be established for the value of k for input integers that follow certain distributions, saving on the fullness of the exhaustive search. However, for many integers in real datasets you have to deal with, the distribution cannot be easily bundled into a convenient approximation of a uniform random variable - for example, have you heard of Benford's law? Notwithstanding, note that close to optimal selection of parameters for the encoding would rarely significantly differ in outcome from optimal selection, in practical implementations.

  • Thanks! I do have plenty of test data. I was just thinking that there is some explicit formula how I could get parameter out of statistical momenta or something like that. So instead I just have to try out different values and iterate to the best? – Džuris May 19 '14 at 10:26
  • And also - I know that it's common and easy to take $M=2^k$, but, as I asked in the question - why would I ever want to choose other $M$ if anything between $M=2^k$ and $M=2^{k-1}$ takes up at least as much bits as $M=2^k$? – Džuris May 19 '14 at 10:28
  • Yes to your first question. To the second, the encoding is a mathematical formulation that is separate from the abstraction of bits and the computation machine. Unlike in theory, in practice you of course have to use a round number of bits. – mockinterface May 19 '14 at 10:47
  • And if I believe that, for example, the data follow geometric distribution, I could find p=1/E(x) and get that M=-1/log_2p or even k=-log_2(-log_2p) if I got it right...? – Džuris May 19 '14 at 10:53
  • @Juris Sorry but I did not get it: If I know that my dataset is geometrically distributed and have E(x), can I choose M to be next power of 2 after E? – ponkin Dec 14 '16 at 8:11

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.