Juice v0.4 refactor

[khosravipasha]

• [ ] Param Learning

  • [x] EM
  • [ ] EM CPU
  • [ ] Regurlarization (Softness)
  • [ ] SGD version. Best option seems to be to customize rrules from ChainRules.jl. Then can grad/optimize easily with Zygote or Flux.

• [ ] Queries/Transformation for the new BitCircuit type. CPU and CUDA versions.

  • [x] EVI
  • [x] MAR
  • [x] MAP (approx)
  • [x] Sample
  • [ ] Conditionals
  • [ ] All marginals
  • [ ] Marginal MAP algorithm. Would need to rethink the algorithm a bit since now have any input types.

• [ ] Structures

  • [ ] Splits, and clones
  • [ ] Learn structures from missing data
  • [x] HCLT
  • [ ] Strudel
  • [ ] RAT-SPN with option to specify leaf type
  • [ ] PD Structures for images?
  • [ ] Other common structures (fully factorized, NB, HMM, markov chains, etc)

• [ ] Tests

  • [ ] Tests for EM learning with missing in the inputs
  • [ ] Test with different input types for likelihood, flow, MAP, Sample, etc
    • [ ] Binomial
    • [ ] Categorical
    • [x] Indicator
  • [ ] Better small circuits for test cases, little_4var might be too simple.
  • [ ] Test for conditional sampling of circuits

• [x] Docs

  • [x] GPU Docs in manual
  • [x] Notebook for learning
  • [ ] This seems nice for adding citations if needed https://github.com/ali-ramadhan/DocumenterCitations.jl

• [x] Misc

  • [ ] #120
  • [x] Float input types had some issues, need to double check
  • [x] Lazy Batching (batchview??) not sure if worth it https://mldatapatternjl.readthedocs.io/en/latest/documentation/dataview.html#as-vector-of-batches

• [x] IO

  • [x] DensityEstimationData.jl
  • [x] JuiceModelZoo.jl
  • [x] Read/Write of circuits text format.
    • [x] Faster Version
    • [x] Support direct read/write to zipped format gzip, etc
  • [x] Binary format for circuit for more efficient saving
  • [x] Zoo related IO; zoo_psdd, zoo_pc, etc.
  • [x] TikZ plotting now broken, need slight modification