Variant of `codify` that also returns the outcomes.

[kahaaga]

Currently, codify applies some discretisation to an input vector of observation which encodes data points in a sliding-window-manner (e.g. over embedding points constructed from the input data). Each data point is encoded into an integer. It would be nice to have a function that also returns the outcomes corresponding to the integer.

This would allow much nicer Counts/Probabilities upstream:

using Associations
x, y = StateSpaceSet(rand(1000, 2)), StateSpaceSet(rand(1000, 3))

 # min/max of the `rand` call is 0 and 1
precise = true # precise bin edges
r = range(0, 1; length = 3)
binning = FixedRectangularBinning(r, dimension(x), precise)
encoding_x = RectangularBinEncoding(binning, x)
encoding_y = CombinationEncoding(RelativeMeanEncoding(0.0, 1, n = 2), OrdinalPatternEncoding(3))
discretization = CodifyPoints(encoding_x, encoding_y)

# now estimate probabilities
probabilities(discretization, x, y)

julia> probabilities(discretization, x, y)
 4×12 Probabilities{Float64,2}
       6      8      12      11      9      1      10      2      4      7      3      5
 1     0.024  0.022   0.03    0.019  0.03   0.029   0.018  0.02   0.023  0.025  0.023  0.022
 4     0.023  0.027   0.017   0.019  0.022  0.022   0.016  0.019  0.029  0.014  0.017  0.014
 3     0.021  0.019   0.015   0.018  0.021  0.028   0.021  0.018  0.016  0.021  0.027  0.019
 2     0.013  0.033   0.016   0.024  0.019  0.02    0.017  0.022  0.018  0.02   0.01   0.02

It would be nice to know what these integers in the marginals correspond to.

This isn't by any means necessary, but it would be a nice addition.

[Datseris]

the only way this would be non breaking is with a wrapper function such as codify_into_outcomes that calls codify and then does the extra step.

[kahaaga]

the only way this would be non breaking is with a wrapper function such as codify_into_outcomes that calls codify and then does the extra step.

Yep, an additional function is the way to go.