Skip to content
Started 5 yr 6 mo ago
Took 43 min on reynolds
Success

#48 (Sep 23, 2018, 5:38:55 PM)

Changes
  1. get some initial code thoughts up.. (details / bitbucketweb)
  2. quadrature params stuff (details / bitbucketweb)
  3. personal add to gitignore (details / bitbucketweb)
  4. merge in branch (details / bitbucketweb)
  5. minor (details / bitbucketweb)
  6. done plots for BOD (details / bitbucketweb)
  7. working on ice model (details / bitbucketweb)
  8. minor testing (details / bitbucketweb)
  9. working on multi-dimensional ice model (details / bitbucketweb)
  10. stopping crit based on first iteration. (details / bitbucketweb)
  11. pre_solve and post_solve methods (details / bitbucketweb)
  12. working on adaptivity for spatially varying ice model (details / bitbucketweb)
  13. push for daniele (details / bitbucketweb)
  14. modified interface for refinement (details / bitbucketweb)
  15. modified refinement for sparse grid (details / bitbucketweb)
  16. initialized map to identity prior to minimize_kl_divergence first call (details / bitbucketweb)
  17. move sparse grid adaptation (forming the sparse grid)  into 'quadrature' (details / bitbucketweb)
  18. current tests on BananPlots (details / bitbucketweb)
  19. add comments for myself for tomorrow (details / bitbucketweb)
  20. working on ice settin (details / bitbucketweb)
  21. 'GradientDimensionAdaptiveSparseGridKLMinimizationValidator' (details / bitbucketweb)
  22. working on ice model (details / bitbucketweb)
  23. working on whitening of ice model (details / bitbucketweb)
  24. working on ice (details / bitbucketweb)
  25. better inverse map for ice problem (details / bitbucketweb)
  26. minor (details / bitbucketweb)
  27. sparse grid is not extracting f from kwargs, to be passed every time (details / bitbucketweb)
  28. preparing for run (details / bitbucketweb)
  29. trying modification to variance diag calculation (details / bitbucketweb)
  30. try except in _refinement of KLAdaptivity (details / bitbucketweb)
  31. sign error. BOD (details / bitbucketweb)
  32. working on ice (details / bitbucketweb)
  33. modified external interface for (x,w) for sparse grid library. (details / bitbucketweb)
  34. BananaPlots works perfectly. (details / bitbucketweb)
  35. BananaPlots MC vs SG. curently the MC doesnt work because in my branch (details / bitbucketweb)
  36. just some printing for diagnosing (details / bitbucketweb)
  37. consistent naming (details / bitbucketweb)
  38. bod plots for mcqmc (details / bitbucketweb)
  39. add the points on the top of the plots for BOD sparse grid (details / bitbucketweb)
  40. kwargs (details / bitbucketweb)
  41. Is this a bug? (details / bitbucketweb)
  42. setting results on paper (details / bitbucketweb)
  43. add arguments to pre/post solve (details / bitbucketweb)
  44. testing ideas... (details / bitbucketweb)
  45. add x,w, to dictionary after error estimation (details / bitbucketweb)
  46. validators (details / bitbucketweb)
  47. finished with tests for this branch. starting new branch for quadrature (details / bitbucketweb)
  48. pushing commented out code that shows the error of log(0) (details / bitbucketweb)
  49. changed tolerance for bod problem (details / bitbucketweb)
  50. figuring out tolerances (details / bitbucketweb)
  51. Remove class GradientDimensionAdaptiveSparseGridKLMinimizationValidator. (details / bitbucketweb)
  52. comments, and justfication for changes with comments  for Daniele (details / bitbucketweb)
  53. remove some unnecesssary code. (details / bitbucketweb)
  54. remove more comments, etc (details / bitbucketweb)
  55. testing default_initial_quadrature. (details / bitbucketweb)
  56. Trying to case on whether or not 'initial' grid was laready created. (details / bitbucketweb)
  57. some order in the tolerances (maybe) (details / bitbucketweb)
  58. sort weights by absolute value for more stability in order of addition. (details / bitbucketweb)
  59. Place plotting under a boolean for toggling ease. (details / bitbucketweb)
  60. commenting and refining quadrature... (details / bitbucketweb)
  61. clean up quadrature qtype 4 (details / bitbucketweb)
  62. remove 'np.newaxis' everywhere. use [x], now that I understand the (details / bitbucketweb)
  63. no need for 'previous_fval' under current algorithm.. (details / bitbucketweb)
  64. KeyError exception for default quadrature (details / bitbucketweb)
  65. KeyError for presolve. (details / bitbucketweb)
  66. coment an issue with this branch. why is pull_tar.n_coeffs missing if we (details / bitbucketweb)
  67. add ref_params to post_solve (details / bitbucketweb)
  68. add ref_params to post_solve, logger info (details / bitbucketweb)
  69. Fix BananaPlots for current interface (details / bitbucketweb)
  70. Oops, I accidentally merged branches incorrectly. Just rolling back some (details / bitbucketweb)
  71. add stochastic volatility to this branch. (details / bitbucketweb)
  72. trying much simpler interface to functions/integrands for sparse grid (details / bitbucketweb)
  73. bug fix (details / bitbucketweb)
  74. space (details / bitbucketweb)
  75. comments (details / bitbucketweb)
  76. add stochastic volatility  to this branch (details / bitbucketweb)
  77. change tol for StochasticVolatility.. decrase max_grid_pts to alleviate (details / bitbucketweb)
  78. rethinking instantiation of parameters for GaussianDistribution (details / bitbucketweb)
  79. reshaping the quadrature function in GaussianDistribution (details / bitbucketweb)
  80. using bfgs in adaptivity for now to make sense of tolerances. working on (details / bitbucketweb)
  81. we don't need 'converged', or 'termination reason' or the integral (details / bitbucketweb)
  82. BFGS for the BOD problem (details / bitbucketweb)
  83. now adaptive_quadratures can be built also wrt for pushforward/pullback (details / bitbucketweb)
  84. relegate checking for uniqueness if x,w to some 'debug' mode in future (details / bitbucketweb)
  85. start testing laplace approximation. new BOD file for that. (details / bitbucketweb)
  86. added missing file (details / bitbucketweb)
  87. mirroring what we've done in frozendistributions quadature, dont need to (details / bitbucketweb)
  88. adapter -> qparams['adapter'] (details / bitbucketweb)
  89. techincally dont need two files, just separate them though so we can (details / bitbucketweb)
  90. nax = np.newaxis (details / bitbucketweb)
  91. standard normal doesn't need trnasformation (dont need to use sampling (details / bitbucketweb)
  92. right now quadratur requires abs_tol, even if initial grid is just an (details / bitbucketweb)
  93. pre/post solve fixes try except - > if else (details / bitbucketweb)
  94. pre/post solve fixes (details / bitbucketweb)
  95. Decrease optimizaiton tol for the Laplace Approx problem, since its (details / bitbucketweb)
  96. testing parameters for initial_quadrature (details / bitbucketweb)
  97. I believe you need to warmstart here. (details / bitbucketweb)
  98. use error estimation points in post_solve (details / bitbucketweb)
  99. allow first variation/line serach to use the last set of points from (details / bitbucketweb)
  100. add L^2 regularization part of objective gradient. this right now is (details / bitbucketweb)
  101. add regularization to the laplace approx example (details / bitbucketweb)
  102. Revert "add L^2 regularization part of objective gradient. this right (details / bitbucketweb)
  103. slash \, and justification for qparams abs_tol (details / bitbucketweb)
  104. add norm printing (details / bitbucketweb)
  105. more informative comment. realized that the regularization is (details / bitbucketweb)
  106. stuff for l2 regularization (details / bitbucketweb)
  107. parameter adjustments: decrease regularizaztion, relax the initial (details / bitbucketweb)
  108. make the optimization a bit easier for this regularized problem. too (details / bitbucketweb)
  109. setting up Poisson (details / bitbucketweb)
  110. almost done with the constant field Poisson setting (details / bitbucketweb)
  111. fixing gradients (details / bitbucketweb)
  112. Poisson problem set up for constant field and Gaussian field (details / bitbucketweb)
  113. whitening script for Poisson (details / bitbucketweb)
  114. added action of the Hessian (details / bitbucketweb)
  115. fixed whitening script and added xml map descriptor (details / bitbucketweb)
  116. fixed bug in FrozenDistributions qtype==0 (details / bitbucketweb)
  117. change how w outputs in darkspark (details / bitbucketweb)
  118. increase max allowed levels... all quadarture levels were maxing out at (details / bitbucketweb)
  119. increase max allowed levels to a very high amount.. (details / bitbucketweb)
  120. fix a bug (details / bitbucketweb)
  121. working on sv (details / bitbucketweb)
  122. working on new active variable detection technique (details / bitbucketweb)
  123. turn try except into if/else. comment out new metho for  prunning of (details / bitbucketweb)
  124. abs_tol is not needed for the default quad case. bring back 'self.eps', (details / bitbucketweb)
  125. if BFGS with inf norm, print inf norm of jacobina, not 2 norm. (details / bitbucketweb)
  126. eps in validator (details / bitbucketweb)
  127. working on SV. Try ½ solve tol to start with. ½ each time. set a minimum (details / bitbucketweb)
  128. prints (details / bitbucketweb)
  129. Informative printing of info, revert to 1/10 solve_tol, seems to work. (details / bitbucketweb)
  130. start MC SV (details / bitbucketweb)
  131. working on sv tests for talk (details / bitbucketweb)
  132. figures generated for Josh talk (details / bitbucketweb)
  133. update poisson example for newer versions of fenics (details / bitbucketweb)
  134. undo bug (details / bitbucketweb)
  135. start with diagonal (details / bitbucketweb)
  136. Fixed Poisson problem set up in SG adaptivity (details / bitbucketweb)
  137. made DARKSparK and optional External package (details / bitbucketweb)
  138. mpi works on the poisson example (details / bitbucketweb)
  139. cleaning up poisson problem to be included to the available (details / bitbucketweb)
  140. working on general setting of PDE problems with FENICS (details / bitbucketweb)
  141. going on... (details / bitbucketweb)
  142. going on.. (details / bitbucketweb)
  143. going on.. (details / bitbucketweb)
  144. almost done with the Poisson setting (details / bitbucketweb)
  145. added LinearTriangularMap for fast regression (nested Cholesky) (details / bitbucketweb)
  146. working on setting for Poisson problem (details / bitbucketweb)
  147. working on Poisson (details / bitbucketweb)
  148. added the bootstrap validator and changed arguments on error_estimation (details / bitbucketweb)
  149. added bootstrap option to tmap-tm (details / bitbucketweb)
  150. working on the bootstrap validator (details / bitbucketweb)
  151. changed the stopping criterion for the bootstrap validator (details / bitbucketweb)
  152. polishing adaptivity (details / bitbucketweb)
  153. fixing coefficients update and cache usage in first variation (details / bitbucketweb)
  154. added better estimation of points multiplier for bootstrap validator (details / bitbucketweb)
  155. working on randomized approximation of Hessian for Laplace approximation (details / bitbucketweb)
  156. randomized_direct_svd works (details / bitbucketweb)
  157. randomized svd is consistent with eigenvalue decomposition (details / bitbucketweb)
  158. low-rank laplace work with outer gradients and low-rank hessian (details / bitbucketweb)
  159. implementing likelihood informed subspace procedure (details / bitbucketweb)
  160. working on Laplace approximation with lis in the head of the square root (details / bitbucketweb)
  161. added option lis on tmap-laplace. LinearTransportMaps and (details / bitbucketweb)

Started by BitBucket push by dabi86 (2 times)

Revision: c7ff7512aa8cdc434d309149547bb094b0a5dea0
Repository: git@bitbucket.org:dabi86/transportmaps-private.git
  • refs/remotes/origin/develop/master
Configuration Matrixgccclang
python2
python3