CHANGES
=======

* upd gitignore
* now packaging with pbr

v0.0.2
------

* jobfiles too
* added \_\_pycache\_\_ to gitig
* more sims
* upd did
* some fixes in the data handling -- logging
* some training
* trying
* looks like value caching working
* rozza rocks
* tmp data to dedicated directory
* looks like ok
* upd gitig
* cont
* basic RB functions like working
* now with setup.py
* done with psf
* zenodo badge
* upd readme
* the raw data
* rmd old file
* pubsoftform
* Add license
* upd gitig
* added README
* preparing for code pub
* added a script for all results
* so much update
* bug fix
* gonna log E(yy) too
* scripting the plot diffs
* scripting the pce pod diff plot
* get the vecs the right dimension
* fix
* and do the diff plot
* some more verbosity
* going plot diffs
* now seeding for the mc-basis
* beauti
* now comp only E(yy) -- and the variance later
* testing postprocess
* going variances
* repro pce[2] results
* fixed a naming -- mgp==1.0.1 seems to work
* upd did
* lil sims well scripted
* towards checking variances
* test the pce vs the mc
* done cleanup
* clean up repo
* upds
* pp and need another run
* mrgd
* woe
* comp things
* many files
* post processing
* major error
* post process script
* all there for ppc
* now pod results independent of ref vals
* data collection for mcpod and mc-basis
* now also all poddim data
* added pid for mc sims cached data
* more informative json str
* now with timing runs and json storing the results and timings
* mc tested
* now saving to file rather than multiprocessing.Queue
* timing but cannot queue too large objects
* now multiproc for mc tooo
* more evenly distributed iterations per process -- timings
* can now script any number of processes
* back to normal sh
* now with multiproc for pce -- stil hardcoded
* now also scripting the nus
* now ref vals by now
* testing the convergence
* clean the repo
* mrg fixs
* mrgd
* new meshes
* done and restruc (now with structure)
* looking good -- need control lines for the inner boundary too
* 2 segs go
* now scripting poddims too
* trying to have fine mesh only at the corners
* now also scripting pcedims
* less visc more vari
* new meshes -- refining at the corners
* other range -- fix for nonzero left boundary
* simming
* script+=mcpod, now reporting training pcexpy
* now conv in z direc
* major bug fix
* now aborting if no comp vals
* mrgd
* fixes
* fixes
* minor upds
* more scripting
* typo fixes
* more scripting
* scripting
* scripting
* scripting
* other refinement -- better convergence
* output info
* now with cline options
* need to do go with rhs -- refined observervation domain
* with the boundary -- complete failure
* redid the plotting -- pod with rhs only works kinda
* trying
* not working -- gonna try zero Dirichlets and rhs
* 3D sim like working
* 3D problem setup
* starting the cylinder problem
* proper naming
* mesh works -- now solving
* maybe extrusion is a bad idea -- gonna try without
* dbgng
* updates
* make the 3D mesh please
* just extruded
* some-sims
* upd
* pypath
* rename
* loooking good
* nice results -- no need for syncing >?
* restruc done -- need to check the red mod
* clean up
* redo
* simulating
* looks all good -- need to do statistics of mc-pod -- pce-pod with lower error
* fwd pce like working -- now genpod
* the new module
* on n on
* started refac of pce
* dolfin la backend
* trying around
* restructs
* looks like good -- gonna do a refac
* pod vecs
* multidim pce works
* mc for several uncertain segs working
* mc vs pce
* monte carlo for estimated value
* no outsrcd like problem setup
* can now solve a single prob
* sanity checks pass
* the mesh and mesh functions
