Package CHEM :: Package datatype :: Module compress

Module compress

Classes

Codec

EliasGammaCodec
Encode/Decode integer according to: http://www.inference.phy.cam.ac.uk/mackay/itprnn/code/c/compress/ Integer must be >=1, unless codec is initialized with the "signed=True" FORMAT: -run of zeros of length = 1 - bits needed to encode data -data, first bit MUST be 1 -(optional) 1 bit for the sign notice, that zero cannot be encoded by this scheme.

IntegerCodec
OPTIONS:...

QRCodec

GolombCodec
FORMAT:...

ReorderBits

RandomReorderBits

MEGICodec
FORMAT:...

PowerLawCodec
FORMAT:...

MEGRCodec
MEG vector encoding zero runs.

Functions

[hide private]

log_factorial(N, base=None)
Use the Gosper approximation of a factorial (much better than Stirling) to caculate the log_{base} N!

IntegerWidth(N)

FullInteger(N)

RandomSet(A, N=sys.maxint)

RandomDict(A, N=sys.maxint, maxC=15)

test()

main()

Function Details

[hide private]

log_factorial(N, base=None)

Use the Gosper approximation of a factorial (much better than Stirling) to caculate the log_{base} N!

$log N! \approx 0.5 \log [\pi (2*N + 1/3)]+N \log N - N$