Summary
- New stoc vol file (details)
- Done with implementation stochastic volatility with no hyper-parameters (details)
- Stochastic volatility (details)
- answers to Alessio comments (details)
- started with solver (details)
- stoc vol works. TODO: implement diagnostics (details)
- started with total order implementation (details)
- working on MixedOrder class. (details)
- working on mixed order (details)
- working on MixedOrder (details)
- Added diagnostic routines (details)
- merged into toolbox (details)
- working on stocVol example (details)
- stocVol (details)
- stocVol (details)
- added matplotlib dependency (details)
- working on banana example (details)
- documenting banana (details)
- working on banana (details)
- working on banana example (details)
- banana example needs documentation regarding diagnostics (details)
- added some diagnostics (details)
- working on laplace_approximation function (details)
- laplace (details)
- working on optimizer (details)
- Laplace approx done. To be tested. (details)
- working on FrozenGaussian with precision input (details)
- tests passed for new version of GaussianDensity. TODO: add unittests (details)
- done precision tests (details)
- laplace works. to finish tests (details)
- laplace works. to finish tests (details)
- laplace works. to finish tests (details)
- Laplace ready (details)
- test optimizer (details)
- working on optimizer (details)
- fixed optimizer (details)
- optimizer works! (details)
- added some docs on banana (details)
- added some logging in the Diagnostic module. To be added all over. (details)
- newton_cglu_optimizer (details)
- added function niceplot (details)
- Added class TransportMap (details)
- class TransportMap integrated (details)
- Init maps (details)
- working on the CommonBase LinearSpan TM (details)
- Removed duplicate pullback functions (details)
- working on CommonBase LinearSpan Map (details)
- linear transport map (details)
- fixed naming of TransportMaps (details)
- working on CommonBase LinearSpan Map (details)
- working on CommonBase LinearSpan Map (details)
- Done CommonBasis LinearSpan Map. Moving to (details)
- working on CommonBasisLinearSpanApproximation (details)
- working on CommonBasisLinearSpanApproximation (details)
- working on CommBase Map (details)
- moving function evaluation and gradients into LineaSpanAppoximation (details)
- Done with linear map class. Now moving to unit test (details)
- in process of migrating functions from FullLinearSpan to LinearSpan with (details)
- moved most of the stuff to LinearSpan (details)
- moved most of the stuff to LinearSpan (details)
- LinearSpanApproximations ready. Move to TriangularTransportMap (details)
- Minor changes (details)
- changed uml (details)
- Added plotAlignedSliceMap (details)
- Added niceContour, plot2density and testBananaPullback (details)
- added testBananaPullback (details)
- fixed l2 regularization (details)
- rotated banana improved (details)
- working on #36. Done with Transport maps? (details)
- Banana example linear span (details)
- some mods on the examples (details)
- #26 working on TM densities test cases (details)
- working on test cases (details)
- working on test cases (details)
- working on density test cases (details)
- working on test cases for densities (details)
- working on dens test cases (details)
- working on tests densities (details)
- transport map densities tests added. Still some problems on the linear (details)
- added tests for kl_divergence (details)
- removed old tests structure (details)
- #26 fixed all unittests. TODO add test for LinearTransportMap (Alessio) (details)
- updating UML graph and TransportMap classes (details)
- Fixes #27: UML graph updated (details)
- fixed import name for CommonBasisLinearSpanTriangularTransportMap (details)
- relaxed target eps in MC tests for density unittests (details)
- Jenkins test develop script added (details)
- fixed test_develop chmod (details)
- #25 restructuring precomp strategy in order to enrich the precomp (details)
- #25 restructuring precomp strategy in order to enrich the precomp (details)
- #25 restructuring precomp strategy in order to enrich the precomp (details)
- uni-variate preallocation works for linear system (details)
- working on TransportMapBase to allow both uni-variate and multi-variate (details)
- working on TransportMapBase to allow both uni-variate and multi-variate (details)
- #25 linear span transport maps with precomputation of uni-variate and (details)
- working on IntegratedExp approximation (details)
- working on IntegratedExp approximation (details)
- working on IntegratedExp approximation (details)
- working on precomputation in integrated exponential. Some tests fail... (details)
- #25 preomputation on integrated exponential works! Make comprehensive (details)
- fixed FrozenFunction precomp functions (details)
- KLdivergence example with options (details)
- working on common basis (details)
- working on common basis (details)
- fixing common basis (details)
- CommonBasisLinearSpan ready. TODO: unittests (details)
- making unittests for CommonBasisLinearSpan (details)
- fixed stoc Vol with constant term (details)
- CommonBasis LinearSpan works. Moving to IntegratedExponential (details)
- CommonBasisIntExp started. precomp_evaluate done (details)
- IntExpFunctionApproximation precomputation structure refactorization (details)
- refactored IntExpFuncApproximation (details)
- working on CommonBasis IntegExp (details)
- CommonBasis IntegratedExponential works (details)
- KLdivergence example now uses CommonBasis implementation (details)
- Fixes #25. (details)
- updated banana example with common basis (details)
- TotalOrder approximation added. Fix the integration into IntExp (details)
- Total order working on test_transportmaps. TODO: Add (details)
- Total Order unittests done (details)
- fixed CommonBasis precomputation, adding to_compute_flag (details)
- Fixes #1. Total order truncation implemented. Updated KLdivergence (details)
- Added BOD model and ConditionedDiffusion folders (details)
- fixed conflict on testBananaPullback (details)
- started with new branch (details)
- fixed bug on GaussianDensity rvs (details)
- fixed bug on GaussianDensity rvs (details)
- working on stochastic volatility example (details)
- Added joint BOD model (details)
- reshaping StochasticVolatility example (details)
- BOD model works. (details)
- Stochastic volatility issuing command changed (details)
- #31 started optional vectorization of expectations in (details)
- Notes composition map Stoc Vol (details)
- Added banana folder (details)
- starting with stoc vol composition (details)
- split map generation from StochasticVolatility files (details)
- Added gradients composition (details)
- Added gradients composition (details)
- Added Markov Component Stoc Vol (details)
- stoc vol composition almost there.... (details)
- stoc vol composition runs. TODO: compose all the maps (details)
- working on lifting and overall pullback (details)
- stochastic volatility composition works!!! (details)
- refactoring the vectorization in several routines... to be finished. (details)
- #31 refactoring to allow computations in batches (details)
- refactored everything to execute in batches defined by a list (iterable) (details)
- new implementation works on old unit tests (details)
- current implementation works on unittests and KLdivergence example test (details)
- setting up BOD model for higher dimension... (details)
- Opens #20. Added logging capabilities in the minimize_kl_divergence (details)
- working on BOD higher order (details)
- working on bod (details)
- working on KL divergence arctan function (details)
- cleaned KL divergence file (details)
- re-organized predefined densities (details)
- added corresponding changes also in the LinearTransportMap class (details)
- working on the tilted banana (details)
- updating LinearTransportMap (details)
- updating LinearTransportMap (details)
- working on banana (details)
- added default classes for Isotropic transport maps (details)
- added default classes for Isotropic transport maps (details)
- using default functions all over (details)
- working on rotated banana (details)
- #35 working on cached evaluations low-level (details)
- #35 low-level caching for function evaluation implemented. TODO: check (details)
- removed some all-equal statements (details)
- changed handling of indices (details)
- caching low level partial_xd done (details)
- changed array_equal withing function evaluation with a same_coeffs_flag (details)
- fixed FrozenTriangularTransportMaps (details)
- Gradients conditional diffusion (details)
- working on conditioned diffusion script. fixed inputs to StocVol (details)
- conditioned diffusion running script ready (details)
- using default TransportMap generation all over the place (details)
- Implemented gradients Markov Component Cond Diff (details)
- working on fixing stochastic volatility components. TODO: Fix the (details)
- fixed full transport map with extra same_coeffs_flag (details)
- Debugging Hessian CondDiff (details)
- Fixed gradient Markov Component Conditioned Diffusion (details)
- Added composition maps for conditione diffusion (details)
- working on conditioned diffusion (details)
- FullTransportMap does not need to redefine pullback and pushforward (details)
- cond diffusion works. better parametrization is needed (RBF?) (details)
- can recover precision matrix in 3d (details)
- updating UML diagram (details)
- map from samples now has a couple more distributions to sample from (details)
- mapFromSamples also plots histograms of pullback against original (details)
- removed "TranportMaps." tag from index (details)
- compare map to analytical (details)
- histogram plots a bit nicer (details)
- working on using RBF in banana example (details)
- preparing for high-level caching. Restructured precomp structure (details)
- high-level caching is done. TODO: check it is doing what is required to (details)
- the high-level caching and the computation in batches/parallel need to (details)
- mfs looking good, need support map still (details)
- updated interface min_kl_div in TransportMapBase (details)
- removed low-level caching. added high-level caching. TODO: check that (details)
- added test case for high-level caching. change idxs_list to idxs_slice (details)
- test cases work. try it out on examples and then test performaces on BOD (details)
- works on examples kldivergence. Check that caching is actually used (details)
- #35 high-level caching successful. TODO: profile on the BOD example. (details)
- turned pushforward to exp(log_pushforward) (details)
- experimenting on the banana example (details)
- fixed bug in TransportMapBase (details)
- removed flag vectorize. batch_size determines whether vectorize or not (details)
- updated examples without vectorize flags (details)
- fixed kl_divergence routines with batches. list -> slice (details)
- working on low-level caching (details)
- working on low-level caching (details)
- low-level caching works on test cases (details)
- trying to improve the hess computation (details)
- alternative solution (details)
- added grad_a_squared with precomputed VVT (details)
- removed same_coeffs_flag and added caching for partial_xd (details)
- working on the banana (details)
- reverted to grad_a instead of grad_a_squared (details)
- polishing hess_a in IntegExp (details)
- polishing hess_a (details)
- polishing hessian little by little (details)
- mfsSandia includes transform (details)
- some improvements in hess_a_log_pullback (details)
- working on banana (details)
- fixed import to numpy.linalg (details)
- fixed hessian with regularization (details)
- adding missing files to conditioned diffusion (details)
- working on stoch diff (details)
- working on conditioned diffusion (details)
- improving visualization in conditioned diffusion (details)
- plot in place conditioned diffusion (details)
- working on value at risk problem (details)
- working on the banana (details)
- working on banana (details)
- adding missing files to conditioned diffusion (details)
- working on stoch diff (details)
- working on conditioned diffusion (details)
- improving visualization in conditioned diffusion (details)
- plot in place conditioned diffusion (details)
- working on value at risk problem (details)
- remove rare events (details)
- starting the sparsity algorithm (details)
- sandia samples: now asks whether right or left-tailed (details)
- plot samples includes 1d analytic form (details)
- sandia1d computes analytical pdf and kl between that and computed (details)
- kl div in 1d needs approximation help (details)
- pullback density in progress (details)
- fixed pullback (details)
- trying MSE ? instead of KL (details)
- mse computation in development (details)
- Open #8 (details)
- added derivative checks tests for pushforward transport map densities. (details)
- building up example for inverse construction (details)
- set up kl using TM density (details)
- examples work derivative free (details)
- creating inverse gradients (details)
- working on gradient (details)
- checking gradient (details)
- back to kl (details)
- gradient inverse map works (details)
- implemented gradient. to be checked with tests and stochastic volatility (details)
- grad_a_log_pushforward passes the unit test. Check signs and fix (details)
- grad_a_log_pushforward passes the unit test. Check signs and fix (details)
- unit tests passed. Remove precomputation in minimization in order to (details)
- inverse solution works. todo: profile to see whether some computations (details)
- partition densities off (details)
- working on banana example (details)
- added no-plot option (details)
- Open #19: written test functions for map from samples. The problem has (details)
- maps from samples tests work fine (details)
- working on composite and inverse transport maps, in order to do (details)
- working on the missing composition functions (details)
- adding missing functions (details)
- added grad_x and hess_x functions to Composite maps. (details)
- working on the banana example. transform the target to the pushforward (details)
- adding necessary inverse functions for pushforward (details)
- fixing banana with composition of maps (details)
- updated UML diagram (details)
- banana example works with composite densities (details)
- mfs tests has laplace, needs reorg (details)
- test cases using randomly generated coefficients work (details)
- started with direct construction of the inverse sv transport. Added (details)
- added default also for linear span approximations (details)
- working on sv inverse transport (details)
- woring on stoc vol (details)
- working on cond diffustion (details)
- small changes in defaluts (details)
- Default option poly and rbf (details)
- added rbf/poly option to cond diffusion (details)
- working on storage and reload of computed maps (details)
- construction and postprocessing scripts for conditioned diffusion are (details)
- fixed random conditionals plotting problem (details)
- storage (using dill) and post processing of conditioned diffusion (details)
- test numbers 1-4 in extra densities (details)
- working on plotting of conditioned diffusion (details)
- fixed bug in plotAlignedConditionals (details)
- plotting (details)
- forcing full span when rbf is chosen in default constructors (details)
- condiff: fixed input span to (details)
- testing higher order for cond diff (details)
- fixed problem in aligned map plotting (details)
- started with parallelization of diagnostics (details)
- parallelization of the diagnostics (details)
- cond diff: adaptation (details)
- cond diff: adaptation works (details)
- parallelization works for aligned conditionals (details)
- working on cond diff example and figures (details)
- plot with log log (details)
- plotting diagnostic parallelized with mpi_map v1.0.3 (details)
- updated conditioned diffusion postprocessing files (details)
- added variance based diagnostic (details)
- allow for restart in cond diff example (details)
- working on restarting of cond diff (details)
- plotting diagnostic now stack all the evaluation points together in (details)
- added nprocs in adapted composition cond diff (details)
- fixed bug in slice map plotting (details)
- cond diff with restart works (details)
- fixed problem in inverse which would cause loss of precision. cond diff (details)
- started with the parallelization of variance_approx_kl. Changes are (details)
- added Weibull density to the available test distributions (details)
- moved all pc_density analytical stuff to ExtraDensities (details)
- started with issue #41. made a sketch c++ module (details)
- handling arrays Numpy/C done (details)
- started with weird sampler (details)
- sampler done (details)
- working on construction from samples (details)
- working on Laplace for cond diff (details)
- working on plots cond diff (details)
- some plotting progresses (details)
- example for xun ready. Todo: document it.. (details)
- added readme to the CPP-module example (details)
- removed unnecessary files (details)
- prepared figures for YMM presentation (details)
- working on the direct approximation of the Weibull distribution (details)
- log_pdf in extra doesn't work with algorithm, but algorithm is running (details)
- working on map composition direct transport (details)
- Weibull works. Added asymptotic behavior on the Logistic. (details)
- starting with stochastic volatility (details)
- working on step by step construction (details)
- reordering in progress (details)
- renamed files (details)
- moving in the dark for Logistic and Banana examples... (details)
- added installation tips in the documentation (details)
- reorganizing example file organization (details)
- fixed bug due to new version of SciPy rootfinding routine. tolerances (details)
- rewriting the density (details)
- sv density and data generation ready (details)
- sv: Laplace done (details)
- sv: composition of maps done, with postprocess (details)
- sv: direct map done (details)
- direct map in low dimension (details)
- sv: inverse map construction (details)
- algorithm working, with unnormalized data (details)
- normalizing data, in progress (details)
- sv working on the inverse map (details)
- algorithm with renormalized data (details)
- working on sv inverse (details)
- added grad check on minimiziation. working on sv inverse... (details)
- fixed bug on grad_a_inverse pushforward... (details)
- sv inverse map preconditioned throught laplace pullback (details)
- preparing dataset for Shwab (details)
- algorithm tau change (details)
- updated jenkins links (details)
- reversed to v1.0.0 for initial releases (details)
- data loading script for Robert done (details)
- data loading for Robert works (details)
- Added example with only the material for Shwab (details)
- added test pullback for shwab material (details)
- mc study of map parameters (details)
- memory leak issue (details)
- renamed variables (details)
- sv: inverse map preconditioned with diagonal Laplace (details)
- debugging memory leak (details)
- added if conditions around logging outputs (details)
- partially solved memory leak. There is still an issue with precomputed (details)
- improved logging (details)
- New stoc vol with hyperparameters doc (details)
- fixing memory leak (details)
- Done with the derivation of the Stoc Vol model with Hyper-params (details)
- cleaned msf_kl_test file (details)
- mc_sparsity.py (details)
- post processing mc study (details)
- diagnostics are now parallelized (details)
- #45 added default constructors to examples. TODO: fix Gumbel (details)
- Fixes #45 (details)
- finished with default constructors examples (details)
- post proc for mc study has S and Omega matrix and element-wise norms (details)
- Implementing gradients Stoc Vol with Hyper-parameters (details)
- caching passes unit tests. TODO: test on example and check speedup (details)
- added time tracking for examples (details)
- working on examples (details)
- added timing in all the examples (details)
- taking top-down approach. started with kl_divergence functions in (details)
- variable elimination added but has bug, need to find cliques in matrix (details)
- applied changes to all functions in Routines.py. TODO: think about the (details)
- grid graph implemented (details)
- designed a way to make chacing and batching work while parallelizing the (details)
- working on parallel structures (details)
- parallelization of kl_divergence done... maybe. (details)
- working on parallel batching... (details)
- working on parallel-batching... (details)
- MPI implementation to be debugged (old and new unit tests) (details)
- old unit tests work!! (details)
- parallel evaluation of kl divergence work almost on all test cases.. (details)
- parallelization now allows for the use of a pool of processors which are (details)
- added unit tests for minimization. to be tested (details)
- kl tests (details)
- working on unit tests (details)
- fixed test case for arctan parallel (details)
- minimization tests passed (details)
- Done with full joint Stoc Vol with Hypers (details)
- bug fixed. running all tests! (details)
- fixed bugs! MPI works! TODO: update the examples (details)
- Done with Markov Component of Stoc Vol with Hypers (details)
- parallelization works on the examples. TODO: profiling (details)
- fixed bug on unit tests (details)
- speed up unit tests kl divergence (details)
- parallelized precomputation and fixed bug in test_kl_divergence (details)
- another bug in test_kl (details)
- another bug in test_kl (details)
- running tests on stochastic volatility with hyperparameters (details)
- working on parallelization of generation of new samples (details)
- parallelized generation of samples (details)
- post-proc finds error, variance and bias (details)
- fixed mpi_eval call in regression (details)
- fixed bug regression (details)
- removed parallelization in regression for now... opening #52 (details)
- working on composition of maps (details)
- working on stoch vol (details)
- examples work without mpirun (details)
- Fixes #53. (details)
- optional matplotlib in examples (details)
- updated testing script (details)
- made matplotlib optional throughout the module (details)
- added script for running on voyager.mit.edu (details)
- working on stochvol (details)
- working on cluster (details)
- fixed bug in kl divergence unit tests (details)
- working on cluster (details)
- Added trim density on Stoc Vol Hyper (details)
- map generation for density with hyper parameters (details)
- working on postprocessing (details)
- Fixed deepcopy issues with Stoc Vol (details)
- working on cluster (details)
- postprocessing step (details)
- fix typo in Routines.py (details)
- fixed map generation with hyper (details)
- all tests pass (details)
- changes made on hess_x_log_pullback. TODO: replicate on other hess_x (details)
- Issue #31: removed construction of m x d x d x d x d tensor from (details)
- working on #40: implementing the map-reduce scheme (details)
- the map-reduce scheme works. Figure out why the number of processes (details)
- Fixes #53: implements map-reduce scheme (details)
- histograms of map and omega entries (details)
- added the stochastic Laplace postprocessing for the hyperparamters case (details)
- laplace approximation posprocess for hyper parameters (details)
- Implemented Markov component stoc vol hyper (details)
- added defaults for components with different basis and span (details)
- ready to test (details)
- fix bug on defaults (details)
- almost testing (details)
- fixes.. (details)
- something is running (details)
- Scripts composition stoc vol with hyper working (details)
- fixed defaults (details)
- working on unit tests for clusters (details)
- Ready for full problem with 1hyper (details)
- Ready for full problem with 1hyper (details)
- unit tests on cluster work (details)
- Fixes for full problem with 1hyper (details)
- Stoc Vol Hyper--configuration for 2 hyper (details)
- Stoc Vol Hyper--configuration for 2 hyper (details)
- working on first variation examples (details)
- working on example (details)
- working on first variation (details)
- fixed bug on defaults (details)
- other bug.. (details)
- added job number in submission script for clusters (details)
- added voyager submission script (details)
- Setup for 3hypers Stoc Vol (details)
- Setup for 3hypers Stoc Vol (details)
- very unsure on where to put the first variation implementation (details)
- regression example works on python3 (details)
- working on parallelization of regression (details)
- regression is now parallel. TODO: implement map-reduce for regression (details)
- warm start with skipping steps done. It should be tested first! (details)
- Changes to plot and postprocessing (details)
- histograms have no ticks or labels (details)
- added data for 2hyper (details)
- added results for 1hyper (details)
- working on hypersonic. added lag on composition construction. (details)
- Getting rid of the recursion for composition of maps (details)
- Input NMC to post processing script (details)
- first variation implemented. TODO: test it (details)
- added mpi_pool argument to diagnostic functions. Changed design of (details)
- fixed variance diagnostic routine (details)
- fixed routine variance diagnostic (details)
- data 3hyper (details)
- working on refinement (details)
- working on test case 2d (details)
- fixed file TransportMapDensities. The PullBackDensity was defined (details)
- pdf as exp(log_pdf) (details)
- Fixes to post-processing trim script (details)
- working on plots (details)
- working on plots of the first variation (details)
- Decoupled data analysis from plots (details)
- working on plots of first variation 2d (details)
- added diagonal transport map (details)
- adaptivity works on banana (details)
- adaptive refinement works on banana with hyper-cross enrichment (details)
- working on the banana (details)
- done with the banana example (details)
- started with stochastic volatility (details)
- added results 2 hyper (details)
- working on adaptivity for stochastic volatility (details)
- it somehow works on stoc vol 5 steps (details)
- removed data from branch (details)
- new enrichment and stopping strategy in first variation adaptivity. (details)
- preparing sensitivity tests (details)
- refinement routine done. to be tested (details)
- working on full algorithm for adaptivity (details)
- working on remove coefficients (details)
- first variation approach designed (details)
- added input arguments to scripts (details)
- 2hyper map 100 (details)
- refinement strategy includes check on sensitivity of variance diagnostic (details)
- data for stoc vol 2 hypers n100 (details)
- added analysis 2hyper 100steps (details)
- working on figures (details)
- working on plots (details)
- map from samples notebook started (details)
- post processing (details)
- map from samples notebook, code fine but maybe should have more text (details)
- moved samples notebook (details)
- new storage for first variation script (details)
- fixed storage adaptivity (details)
- fixing adaptivity for quadratures (details)
- working on examples for presentation (details)
- added output on refinement (details)
- improved total order multi index routine. (details)
- added output to adaptivity (details)
- getting data.. (details)
- running conditionals for alessio (details)
- improved postprocess (details)
- fixed adaptivity routine (details)
- right now the mpi design is a big mess... find an elegant way to get all (details)
- restructuring all mpi calls according to the new design (details)
- working on parallelization of objective functions (details)
- objective done for integrated exponential... (details)
- modifying regression routines. Todo also: fix the constraints in the (details)
- done with code rewriting. TODO: rewrite all tests and examples. (details)
- debuging: testing serial execution kl_divergence (details)
- testing minimization serial (details)
- minimization IE works. TODO: LS minimization and all parallel (details)
- all serial tests work (details)
- working on parallel tests (details)
- testing parallel on kl_minimization and L2_minimization (details)
- testing parallel on kl_minimization and L2_minimization (details)
- passing all tests (details)
- started with the documentation of MPI (details)
- going on with docs (details)
- fixed pdf/log_pdf for transprot map densities and going on with mpi (details)
- working on example (details)
- working on documentation (details)
- working on documentation (details)
- working on mpi examples (details)
- working on mpi examples.... (details)
- going on with mpi documentation. TODO: check examples... (details)
- updated examples (details)
- examples have been updated to use the new mpi interface (details)
- updated slurm script for examples (details)
- residual modifications to adaptivity routines (details)
- fixed bug in mpi_eval (details)
- working on examples (details)
- updated regression examples to use MPI (details)
- added action of hessian on routines and TransportMapBase. TODO: modify (details)
- action of the Hessian implemented for the kl-divergence and the l2 (details)
- finished last merge (details)
- working on 3hyper (details)
- fixed regression (details)
- fixed chacing of hessian. Now it is really not recomputed (details)
- fixed output hess (details)
- fixed typo in routines (details)
- batching implemented to be exthensively tested (details)
- working on unittests for l2_misfit (details)
- first test passed (details)
- adding tests (details)
- passing some of the tests... (details)
- misfit tests all pass (details)
- regression batching seems to work on unit test cases (details)
- almost ready to run 3hyper on cluster (details)
- almost ready to run 3hyper on cluster (details)
- fixed runner (details)
- removed data. use Dropbox folder for this (details)
- logging capabilities improved for both kl-divergence and regression (details)
- fixed scripts (details)
- fixes #63 (details)
- fixed serialization of the logger (un-pickable and thus removed from the (details)
- added curvature output on Hessian output (to check positive (details)
- mpi tutorial is done (details)
- working on cpp examples (details)
- updated stocvol analysis with new MPI (details)
- updated construction scripts (details)
- updated analysis (details)
- udated plotting for stoc vol (details)
- updated mpi-usage (details)
- re-activated var_diag_list (details)
- added debug code for Hessian conditioning (details)
- added no-marginals option in plotting (details)
- use gaussian graphs in algorithm (details)
- working with generated data (details)
- working on newly generated data (details)
- fixed analysis script (details)
- removed data folder durbin (details)
- updating adaptivity to new MPI (details)
- fixed merge with master. MPI functionalities updated (details)
- added possibility of pulling back through Laplace prior to adapt the map (details)
- improved postprocessing (details)
- added possibiltiy do inverse adaptivity (details)
- working on log-gauss Cox (details)
- added test for processes (details)
- working on log-cox densities (details)
- working on logcox (details)
- working on log-cox (details)
- working on setup of log gaussian cox (details)
- data generation done (details)
- working on log gaussian cox (details)
- plotting change in laplace postprocess (details)
- added pieces of grad_a_omega (details)
- testing new derivatives (details)
- cleaning before merge (details)
- fixes on the full transport maps (details)
- reorganizing examples (details)
- added Laplace scripts (details)
- restructuring examples (details)
- restructuring examples (details)
- restructuring stochastic volatility example (details)
- restructuring stochastic volatility (details)
- restructuring LogGaussCox (details)
- restructuring (details)
- working on LogGaussianCox postprocessing for generic maps (details)
- working on LogGaussianCox (details)
- started working on script for generic transport map construction (details)
- updated contributors list and author list in files (details)
- updated readme (details)
- updated readme (details)
- working on direct map script (details)
- started with xml specification of maps (details)
- delta method almost done, fibers incorrect in TransportMapBase (details)
- working on xml descriptors (details)
- working on xml descriptor parsers (details)
- updated authors and emails (details)
- added xml loader for tests (details)
- reduced dimension of linear map Hessian return (details)
- updated authors and emails... (details)
- working on constructor of linear span approximations (details)
- working on the XML schema... (details)
- working on the schema and validating it... (details)
- working on the definition of a syntax for the components (details)
- working on the definition of a syntax for the components (details)
- working on basis (details)
- added example structures (details)
- working on the parser in TriangularTransportMaps.py (details)
- working on the parser for IntegratedExpTriMap (details)
- updated author and contact in init file (details)
- parser.... (details)
- sprinting on LogGaussianCoxProcess. The script tmap-direct will need to (details)
- test was run successfully for nobs=30. the tmap-direct script is in (details)
- working on independent samplers (details)
- working on rejection sampler (details)
- working on samplers (details)
- working on effective sample size criteria... (details)
- ess implemented (details)
- working on postprocess scripts (details)
- bigger head map for log gaussian cox (details)
- working on the sampling of tmap (details)
- almost done with figures for Log Gauss Cox (details)
- working on Metropolis samples. Added h5py dependency (details)
- tests gradahessxpartialxd (details)
- one bug less to be fixed... (details)
- added test_functions (details)
- figures and scripts for log gauss cox are done (details)
- added batching for tmap-postprocess but needs to be improved (details)
- streamlining scripts for sequential inference postprocessing (details)
- heavy modification of the postprocess scripts for stoch vol. Sequential (details)
- scripts for paper working... (details)
- working on postprocessing plotting (details)
- merging (details)
- prepared plotting for paper (details)
- almost all figures are done (details)
- Christmas plots (details)
- working on figures (details)
- working on parsing of IntExpNonLinearHeadMap-5d.xml (details)
- working on parsing of IntExpNonLinearHeadMap-5d.xml (details)
- working on functional approximations. Allow for the use of (details)
- added the approx layer in order to mimic the class hierarchy (details)
- think a better design for tensorized/non-tensorized basis... (details)
- working on parsing of functionals (details)
- working on tensorization of basis.... (details)
- moving on to the parsing of basis (details)
- constructor works for IntExpNonLinearHeadMap-5d.xml and (details)
- removed HardVars which is cretated at sdist time (details)
- added creation of hardcoded variables (xml schemes) at install and (details)
- working on serializing xml unit tests. (details)
- figure out how to define classes on the fly (details)
- full order works (details)
- monotonic lin span approximation get instantiated, but we need to add a (details)
- tolerance computed with delta method (details)
- working on RadialBasisFunctions support... Maybe we need to interface (details)
- loading from xml works, but multi-index loading and xml storing (details)
- construction of multi-idx map from xml works (details)
- working on tests for scripts (details)
- done with the tests for the scripts (details)
- started working on integrated squared approximation (details)
- working on test_transportmaps inverse.. (details)
- working on kl divergence tests (details)
- working on tests and examples (details)
- test_xml works (details)
- removed Full and Total order approximations. Now they are all (details)
- starting the implementation of the integrated squared linear span (details)
- working on the intsq approximation (details)
- going on... (details)
- progressing... (details)
- we are over the hill.. (details)
- Class IntegratedSquaredParametricFunctionApproximation done. TODO: (details)
- test_transportmaps pass (details)
- working on minimization tests... something wrong with batching.. (details)
- working on int sq (details)
- interfacing with PosDefConstSquaredExtHermitePhysicistsFunction (details)
- something is wrong with the pos def squared implementation... (details)
- working on test_transports for new Inetgrated Squared approximations (details)
- squared const ext hermite prob func works but slow. To be profiled (details)
- integrated squared is almost ready. Implementation sufficient for merge (details)
- ordering and post-processing (details)
- added functions needed for variance of omega, for algebraic basis (details)
- added n_edges to graphs and normalized delta plots (details)
- fixed postprocess script (details)
- fixed input/raw_input (details)
- working on additional functions for list composite map (details)
- working on hess_x (details)
- implemented ListCompositeTransportMap. It's missing some (details)
- imporved documentation tmap-direct (details)
- cleaning: mapsFromSamples dir (details)
- working on scripts interface (details)
- working on help for scripts (details)
- working on scripts (details)
- working on scripts (details)
- algorithm with graphs is now test file (details)
- working on postprocessing scripts (details)
- fixed disclaimer LGPL (details)
- tests work for the scripts in place. start docs. (details)
- restructured documentation. Prepared for scripts documentation (details)
- working on documentation of cli (details)
- working on bod (details)
- fixed references and working on bod (details)
- working on unbiased samplers (details)
- working on unbiased sampling (details)
- data generation for bod (details)
- done with bod (details)
- done with bod (details)
- working on xml docs (details)
- xml docs done (details)
- xml done (details)
- xml done (details)
- gumbel ready for presentation too (details)
- fixed gumbel also for documentation (details)
- added grad_x_inverse and log_deg_grad_x_inverse methods (details)
- working on bod presentation (details)
- going on with bod presentation (details)
- almost done with bod (details)
- bod for presentation ready (details)
- done with tm examples (details)
- done with xml (details)
- started with cli (details)
- started working on the stochastic volatility density... (details)
- started with writing of classes for decomposable densities (details)
- started with writing of classes for decomposable densities (details)
- going on with densities (details)
- working on classes for sequential inference (details)
- working on rewriting stocvol densities (details)
- wrapping up stocvol density (details)
- testing new implementation! (details)
- equivalent density! TODO: MarkovComponent (details)
- working on markov component. (details)
- working on markov comp (details)
- one residual error in the hessian... (details)
- The density is set up and ready to be used (details)
- working on documentation and refactored examples of densities (details)
- documenting stoc volatility (details)
- going on with sequential inference example (details)
- working on docs for stocvol example (details)
- reorg, now with dirs (details)
- fix directory for monte carlo run (details)
- adding data dir (details)
- removed tabulate (details)
- fixed import for python2 (details)
- added missing densities (details)
- removed some example densities for now (details)
- acme data clean (details)
- working on stoc vol example (details)
- working on first step stoc vol (details)
- working on first step stoc vol (details)
- almost done with stoch vol (details)
- doing stoc vol post processing (details)
- sequential inference needs a longer example (details)
- started with SeqHMM integrator (details)
- finished sequential inference... (details)
- fixed imports of integrator. TODO: documentation of the integrator class (details)
- fixed typos in examples (details)
- added svg data assimilation (details)
- added inkscape preamble (details)
- updated figures and notation in sequential inference (details)
- fixed pictures (details)
- fixed bod (details)
- working on filtering (details)
- done with getting filtering out (details)
- ok with filtering of stoc vol (details)
- added the script for sequential inference (details)
- first implementation of sampler works on postprocessing (details)
- added stochastic volatility folder (details)
- adding ess estimation in unbiased sampling script (details)
- postprocess sampling fixed (details)
- working on command line commands (details)
- working on command line documentation! (details)
- working on command line documentation! (details)
- working on cli (details)
- going on with cli (details)
- working on cli postprocess (details)
- working on cli postprocess (details)
- cli done (details)
- dummy file for volatility data (details)
- added missing file (details)
- fixing stile of documentation (details)
- refactored Density->Distribution (details)
- removed unused functions in Diagnostics.Plotting (details)
- checked default constructors. fixes #34 (details)
- added deprecation flag for old sub-module Denisities (details)
- fixed logging of solving stats (details)
- removed default dependency on optional packages (details)
- disabled parallel tests if MPI is not supported (details)
- downgraded documentation to TM 1.0 (details)
- removed circular import in all the submodules (details)
- local merge (details)
- installation goes through on MAC. Going on with documentation (details)
- working on intro (details)
- checking for broken links (details)
- fixing readme (details)
- fixed wrong merge.. (details)
- working on alinged-marginals. Check on quadratures output as well. (details)
- added marginal plotting (details)
- sequential-tm with restart (details)
- tmap-sequential-tm can now re-start. Added verbose options and (details)
- working on postprocess for sequential algorithm (details)
- added postprocess for sequential inference (details)
- done with scripts (details)
- working on documentation of inverse from samples (details)
- working on inverse sparse transport (details)
- added inverse stochastic volatility example (details)
- added figures for cli.ipynb (details)
- working on class diagram (details)
- working on the api (details)
- removed autosummary (details)
- included inheritance graph. Fixes #27 (details)
- fixed dictionary get for python2. Fixes #75 (details)
- updated version 1.0.0b1 and copyright (details)
- updated info (details)
- working on upload script (details)
- working on release testpypi (details)
- modified warning for non-supported features (details)
- fixed linking to SpectralToolbox. Now it is done at first import (details)
- fixed issue in frozen distributions. v1.0.0b4 (details)
- fixed xml linking (details)
- fixed test_xml and test_scripts (details)
- fixed some settings for unit tests failing on kl-divergence due to fd (details)
- refactoring Examples for release (details)
- cleaning up before release (details)
- working on mirroring (details)
- after moving (details)
- working on public mirroring (details)
- working on public mirroring (details)
- added deprecated frozen denisties (details)
- working on mirror (details)
- workin on mirroring (details)
- fixed (maybe) base_density/base_distribution inconsistency (details)
- fixed (again maybe) base_density/base_distribution inconsistency (details)
- mirroring script ready (details)
- fixed readme (details)
- bitbucket-pipelines.yml created online with Bitbucket (details)
- bitbucket-pipelines.yml edited online with Bitbucket (details)
- bitbucket-pipelines.yml edited online with Bitbucket (details)
- bitbucket-pipelines.yml edited online with Bitbucket (details)
- bitbucket-pipelines.yml edited online with Bitbucket (details)
- bitbucket-pipelines.yml edited online with Bitbucket (details)
- bitbucket-pipelines.yml edited online with Bitbucket (details)
- bitbucket-pipelines.yml edited online with Bitbucket (details)
- bitbucket-pipelines.yml edited online with Bitbucket (details)
- bitbucket-pipelines.yml deleted online with Bitbucket (details)
- working on jenkins script (details)
- testing push with jenkins (details)
- updated jenkins script (details)
- fixing pip for long path.. not really needed for tests on bitbucket (details)
- jenkins... (details)
- jenkins... (details)
- jenkins... (details)
- jenkins... (details)
- jenkins... (details)
- jenkins... (details)
- jenkins... (details)
- jenkins... (details)
- jenkins... (details)
- jenkins... (details)
- jenkins... (details)
- Jenkins config: testing test_scripts (details)
- fixed logging of solving stats (details)
- Jenkins tests (details)
- Jenkins... (details)
- Jenkins... (details)
- Jenkins... (details)
- Jenkins... (details)
- Jenkins... (details)
- Jenkins... (details)
- Jenkins... (details)
- Jenkins... (details)
- Jenkins... (details)
- Jenkins: exit(1) if packages failed (details)
- removed cached dir (details)
- Jenkins... working on setup (details)
- jenkins.. (details)
- jenkins.. (details)
- jenkins... (details)
- jenkins... (details)
- jenkins testing all (details)
- updated email (details)
- removed e-mail contact (details)
- updated readme (details)
- updated readme (details)
- updated readme (details)
- reduce verbosity in tests (details)
- updated scripts and logging (details)
- fixed logging (details)
- removed MPI support for macys-Sierra (details)
- fixed log levels.. (details)
- fixed errors in grad_a_hess_x and grad_a_hess_x_log_det_grad_x (details)
- removed tests for grad_a_hess_x_log_det_grad_x. Filed issue #76 (details)
- updated parallelization and fixed failing test in test_kl_divergence (details)
- fixing test_kl_divergence (maybe) (details)
- turning MPI=True on Macys-Sierra (details)
- fixed linear map in IS kl_divergence (details)
- added jenkins tests for testpypi (details)
- modified linear2d test in order to make it more amenable for PushForward (details)
- simplified setup.py (standardized) (details)
- the user is now required to pre-install numpy (details)
- fixed jenkins script (details)
- removed hardvars (details)
- removed hardvars (details)
- removed pushforward kl_div tests (details)
- split serial and parallel tests (details)
- testing serial (details)
- added mpi option (details)
- fixed check on mpi in jenkins (details)
- added lag on mpi tests in order to accommodate for closing and opening (details)
- set unique random seed=1 (details)
- adding badges to the documentation (details)
- disabled MPI on sierra (details)
- relaxed check on l2 minimization tolerance (details)
- updated badges and upload code (details)
- lower tolerance on transportmap_distributions (details)
- fixed namespace and restored tolerance in transportmap_distributions (details)
- updating index (details)
- added verbose output for failed finite difference (details)
- added verbose output for failed finite difference (details)
- set integrated squared to linear for test_transportmap_distributions (details)
- extending readme (details)
- README.rst edited online with Bitbucket (details)
- README.rst edited online with Bitbucket (details)