Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

A more universal entropy calculation method for sampling based inference #151

Merged
merged 11 commits into from
Feb 23, 2021
Merged

A more universal entropy calculation method for sampling based inference #151

Changes from 1 commit
Commits
Show all changes
11 commits
Select commit Hold shift + click to select a range
9da97e6
Avoid unnecessary matrix inversions
semihakbayrak Jan 23, 2021
85e7582
Reverted parametrization back to mean-variance.
ivan-bocharov Jan 28, 2021
f9a8a4b
Allow entropy calculation for smoothing
semihakbayrak Jan 28, 2021
5b49187
Initial tests for entropy calculation and new SampleList features
semihakbayrak Jan 29, 2021
ac77dbf
Renamed variables in SampleList tests.
ivan-bocharov Jan 29, 2021
fcba589
concatenateGaussianMV now can concatenate distributions of different …
ivan-bocharov Jan 29, 2021
8e08400
Fixed a typo in ruleSPNonlinearSInGX definition.
ivan-bocharov Jan 29, 2021
8e8a513
Folded logIntegrand initialization into a conditional branch.
ivan-bocharov Jan 29, 2021
5907b57
Add more comments about Entropy calculation. Test the new entropy cal…
semihakbayrak Feb 12, 2021
43ac935
Changes in convergence criteria of Laplace approximation
semihakbayrak Feb 18, 2021
376127d
Added DataStructures to the list of dependencies.
ivan-bocharov Feb 23, 2021
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
Reverted parametrization back to mean-variance.
  • Loading branch information
@ivan-bocharov
ivan-bocharov committed Jan 28, 2021
commit 85e758245faf0bedec01091aae76af37ee6dcac1
5 changes: 3 additions & 2 deletions src/engines/julia/update_rules/nonlinear_sampling.jl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -94,10 +94,11 @@ function msgSPNonlinearSInGX(g::Function,

# Compute joint belief on in's by gradient ascent
m_in = gradientOptimization(log_joint, d_log_joint, m_fw_in, 0.01)
W_in = -ForwardDiff.jacobian(d_log_joint, m_in)
V_in = cholinv(-ForwardDiff.jacobian(d_log_joint, m_in))

# Marginalize joint belief on in's
(m_inx, W_inx) = marginalizeGaussianMV(variate, m_in, W_in, ds, inx) # Marginalization is overloaded on VariateType V
(m_inx, V_inx) = marginalizeGaussianMV(variate, m_in, V_in, ds, inx) # Marginalization is overloaded on VariateType V
W_inx = cholinv(V_inx) # Convert to canonical statistics
xi_inx = W_inx*m_inx

# Divide marginal on inx by forward message
Expand Down