Summary
- example on Gumbel 1d (details)
- started with the Helmholtz equation (details)
- integrated SING into TransportMaps (details)
- cleaning SING files (details)
- l1 tests and plotting scripts (details)
- working on Helmholtz example from Ghattas' paper (details)
- working on Helmholtz (details)
- starting to work on actual probabilistic model (details)
- renaming some sun_sing files (details)
- updating run_sing files to run with SI.SING (details)
- actually updating to run with SI.SING (ignore last commit) (details)
- testing l1 (details)
- l1 regularization (details)
- Updated regularized map - SING now runs with 2 types of reg and no reg. (details)
- relaxed rademacher, butterfly distribution, and more properties of (details)
- relaxed rademacher uses uniform over (-1,1), not (0,1) (details)
- node ordering fix (details)
- changing parameters in Helmholtz (details)
- added stub solver for EIT problem (details)
- working on surface gradient (details)
- working on CEM (details)
- fixed EIT working on CEM (details)
- plotting scripts for nips (details)
- turned hessian off in sing (details)
- mc studies 1step algs (details)
- More MC Studies for SING (details)
- working on models (details)
- No electrodes model works (details)
- working on discretized model (details)
- working on discretized model (details)
- gmsh mesh generation for discretized test (details)
- working on complete model (details)
- starting to work on adjoint equations (details)
- messed up with no-electrodes model.. (details)
- working on fast sensor function (details)
- fixed sensor expression (details)
- working on Dirichlet to Neumann model (details)
- working on adjoint for no electrode model (details)
- adjoint done (details)
- gradient and Hessian working (details)
- setting up the posterior distribution (details)
- working on likelihood functions (details)
- working on posterior distribution (details)
- storage/loading of distribution works (details)
- working on action of Hessian (details)
- working on action_hess_a for transport map distributions (details)
- working on minimization of kl divergence using action of the Hessian (details)
- working on kl minimization tests (details)
- unit tests are passing. todo: caching in Likelihood to avoid extra (details)
- the distribution object is now pre-broadcasted to the children in (details)
- fixed test_kl_divergence (details)
- working on caching of Likelihood evaluations (details)
- fixed init function (details)
- The action of the Hessian is ready. Fixes #31. Caching issues reserved (details)
- introducing cached and counted decorators (details)
- adding counting output (details)
- caching to be finished in TransportMaps and Maps in general. Then I need (details)
- still working on maps... (details)
- working on distributions (details)
- testing caching. todo: caching in optimization problems (details)
- added caching and counting wherever needed (maybe missing some spots). (details)
- working on caching in Hessian. Todo: update and run unittests (details)
- adding reset counters functionalities (details)
- counting and resetting works. Caching also seems to work. TODO: (details)
- caching does what is supposed to do! (details)
- added manual test for caching and counting (details)
- working on caching minimization routine (details)
- working on unit tests for kl-divergence (details)
- working on test_kl_divergence.py (details)
- fixing an error at a time... (details)
- kl_min tests work! (details)
- fixing all unit tests (details)
- porting MPI implementation to mpi_map v2.0 (details)
- porting to mpi_map v2.0 completed. ToBeTested!! (details)
- mpi_map v2 integrated and passing unit tests (details)
- gathering of ncalls and nevals count at the end of optimization done (details)
- passes unit tests (details)
- now pointing to MPI_Pool_v2 in mpi_map v2.0 (details)
- returning latest cache within log dictionary (details)
- fixed py2/py3 issue with cached decorator (details)
- fixed py2/py3 issue with cached decorator (details)
- fixed py2/py3 issue with cached decorator (details)
- fixing py2/py3 handling of *args and **kwargs (PEP3102) (details)
- working on target distribution (details)
- working on mpi evaluation of target density (details)
- working on constant field data generation and parallel implementation (details)
- working on speeding up function and gradient evaluatiions (details)
- implemented fungrad (details)
- working of action of Hessian (details)
- mpi implementation not working... (details)
- staring solving simple problems.. (details)
- working on constant field problem and parallelization (details)
- working on the problem set up (details)
- data generation for constant field works (details)
- working on adjoint for field (details)
- fixed gradient! (details)
- working on Hessian of EIT problem (details)
- working on hessian (details)
- fixed counter reset in some classes (details)
- fixing regularization (details)
- fixed caching for tuple_grad_a (details)
- fixed some issues with tests and new counting decorators (details)
- action of the Hessian for EIT works. To be integrated in distribution (details)
- eit hessian works with caching of solution and adjoints (details)
- EIT field distribution works with MPI (details)
- added possibility to add real fields by expression (details)
- disabled plotting if not available (details)
- distributions work on cluster (details)
- made EIT solver into a class (details)
- working on validators (details)
- working on saa (details)
- implementing gradient based validators (details)
- validators... (details)
- working on adaptivity algorithm (details)
- working on the adaptivity algorithm (details)
- adding adaptivity options to tmap-tm (details)
- working on launching script tmap-tm (details)
- changing builder classes to imbed parameters and take distributions on (details)
- going through banana example (details)
- working on adaptivity on banana (details)
- adaptivity on banana almost work (details)
- adaptivity works on banana. a lot to do yet... (details)
- improving parallelization (details)
- splitted first variation computation for better parallelization (details)
- adaptivity works on banana without hyper (details)
- working on low-rank approximation of Hessians (details)
- working on laplace approximation for Bayesian inverse problems, where (details)
- working on low-rank approximation of covariance in bayesian problem (details)
- working on low-rank approximation (details)
- added logging information in postprocessing script (details)
- low-rank Laplace approximation works (details)
- applying Laplace to EIT (details)
- working on composition of Likelihood with maps (details)
- working on composition of Likelihood with maps (details)
- composition of functions and map now is possible. working on laplace of (details)
- working on laplace approximation of EIT (details)
- debugging action of Hessian (details)
- checking hessian action.. (details)
- working on finite difference test for Hessian (details)
- checking Hessian for distribution (details)
- fixed withening script. Now action of Hessian works with MPI (details)
- working on EIT adaptivity (details)
- added missing functions to LinearTransportMap. Added unit tests for it (details)
- modified caching of EIT solutions so that the cache is now pickable. (details)
- added verbosity option in adaptivity algorithms and fixed bug in action (details)
- added intermediate storage capability for adaptivity (details)
- added output for lower and upper bounds in saa (details)
- fixed bug in tmap-tm and in test_distribution.py for EIT (details)
- added distributed sampling for MC sampling in minimize_kl_divergence (details)
- fixed sampling for inverse maps from samples (disabled distributed (details)
- fixed distributed sampling for linear span kl minimization (details)
- fixed action Hessian with regularization (details)
- simplified regularization of action hessian regularization (details)
- fixed relative tolerance in validator (details)
- fixed relative tolerance in validator (details)
- remove unnecessary output form EIT (details)
- changed active variable indicator to E[|x|] (details)
- tryna generate non-gaussian data (details)
- working on reload feature for tmap-tm (details)
- reloading done for first variation adaptivity. to be checked in other (details)
- added some output to First variation algorithm (details)
- fixed restarting and grad_x for target in EIT (details)
- added field full_basis_list to function approximations, to be used in (details)
- fixed adaptivity accordingly to previous full_basis_list at the function (details)
- running aerostruct and copula data (details)
- committing data samples for copula data (details)
- fixed first variation computation (details)
- added plot sample for post process in EIT (details)
- added a non-gaussian chain dataset (details)
- added full_basis_list feature to xml map structure loading (details)
- fixed full_basis_list to save in memory allocation (details)
- minor testing datasets changes (details)
- check for product distribution (details)
- removed matplotlib from imports (details)
- scikit-sparse reqs (details)
- changed googleanalytics requirements (details)
- starting jupyter notebook for sing (details)
- looking at copula data inverse covariance (details)
- working on coarser test case (details)
- fixed several bugs in adaptivity and tests (details)
- fixed get state in first variation builder (details)
- added line search options in tmap-tm (details)
- fixed kladaptivity serialization and regression regularization (details)
- fixed gx_gt_kl batched evaluation (details)
- added sing notebook example (details)
- fixing adaptivity here and there.. (details)
- integrated documentation for SING algorithm (details)
- added postprocessing functionalities for adaptivity (details)
- added double circle phantom (details)
- fixed store function in tmap-tm (details)
- added argument for line search delta (details)
- added delta parameter for line search (details)
- fixed documentation SING (details)
- fixing unit tests; (details)
- fixed laplace approximation starting point (details)
- documenting fv adaptivity (details)
- preparing release v2.0b0 (details)
- v2.0b0: ready for unit-testing (details)
- fixed matplotlib dependence (details)
- fixed process_time PyV3 PyV2 (details)
- fixed processtime name (details)
- fixing bugs for release (details)
- working to fix grax_gradt_kl (details)
- fixed grad_x of first variation (details)
- subtract identity_coeffs + l1_reg to component kl (details)
- removed dependency on FileNotFoundError for compatibility with py2 (details)
- fixed compatibility issue with py2 (details)