Compression of random binary data

[Thomas Jollans]

On 2017-10-25 23:22, danceswithnumbers at gmail.com wrote:
> With every transform the entropy changes,

That's only true if the "transform" loses or adds information.

If it loses information, that's lossy compression, which is only useful
in very specific (but also extremely common) circumstances.

If it adds information, that's poetry, not compression.


-- 
Thomas Jollans