search

zenna / CausalDiscovery.jl

9 stars 4 forks source link

prior and conditional program generation based on BSADM paper #10

Closed riadas closed 4 years ago

riadas commented 4 years ago

Could use some a good deal of cleaning up/documentation, but works.

tree1 = generateTree() # generate arbitrary tree (SEM) getExpr(tree1) # converts tree into readable Julia expression tree2 = proposeTree(tree1) # proposes new tree based on old one getExpr(tree2) # display tree getPriorLogProb(tree1) # prior log prob of generating tree1 getConditionalLogProb(tree1, tree2) # log prob conditioned on tree1 of generating tree2

codecov-io commented 4 years ago

Codecov Report

Merging #10 into master will decrease coverage by 98.98%. The diff coverage is 0.00%.

Impacted file tree graph

@@             Coverage Diff             @@
##            master     #10       +/-   ##
===========================================
- Coverage   100.00%   1.01%   -98.99%     
===========================================
  Files            2       4        +2     
  Lines            2     197      +195     
===========================================
  Hits             2       2               
- Misses           0     195      +195
Impacted Files Coverage Δ
src/MCMC.jl/grammar.jl 0.00% <0.00%> (ø)
src/MCMC.jl/model.jl 0.00% <0.00%> (ø)

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 71f18f1...da5b26a. Read the comment docs.