issues
search
jeff-regier
/
Celeste.jl
Scalable inference for a generative model of astronomical images
MIT License
183
stars
28
forks
source link
issues
Newest
Newest
Most commented
Recently updated
Oldest
Least commented
Least recently updated
filter out more perhaps-bogus light sources to avoid load imbalance
#641
jeff-regier
closed
7 years ago
0
fix Optim.Options type over-restriction and add termination_callback test
#640
jrevels
closed
7 years ago
1
GB LLVM Optimization tracker
#639
Keno
closed
7 years ago
0
allows for passing early termination callback to Optim.Options
#638
jrevels
closed
7 years ago
0
Multi-node joint infer with multiple runs
#637
kpamnany
closed
7 years ago
0
Use seeding constructor of MersenneTwister instead of srand
#636
rfourquet
closed
7 years ago
1
automatic batchsize selection joint infer
#635
agnusmaximus
closed
7 years ago
2
removed checkout of dataframes and dataarray in runtests.jl
#634
jeff-regier
closed
7 years ago
0
Make Celeste work with latest version of StaticArrays
#633
andreasnoack
closed
7 years ago
2
Correctly compute proportion of missed stars/gals
#632
gostevehoward
closed
7 years ago
0
Small tweaks/fixes/cleanups while prepping SC17 results
#631
gostevehoward
closed
7 years ago
0
Synthetic.jl image generation, asinh magnitudes, and various fixes
#630
gostevehoward
closed
7 years ago
2
Errors from Julia type inference?
#629
gostevehoward
closed
7 years ago
1
Fix bug + support plots of posterior z-scores
#628
gostevehoward
closed
7 years ago
1
Remove unused variables
#627
andreasnoack
closed
7 years ago
1
Joint infer dynamic
#626
agnusmaximus
closed
7 years ago
1
Tracking issue for SC17 scientific results
#625
gostevehoward
closed
7 years ago
16
Add code to examine errors relative to posterior SDs
#624
gostevehoward
closed
7 years ago
0
get rid of vectorized ops which were causing compatibility issues
#623
gostevehoward
closed
7 years ago
0
Multinode joint inference
#622
kpamnany
closed
7 years ago
0
report the number of active and inactive pixels visited…
#621
jeff-regier
closed
7 years ago
0
Multinode functionality from the `multijoint` branch
#620
kpamnany
closed
7 years ago
3
Script to generate FITS images from a catalog using `Synthetic.jl`
#619
gostevehoward
closed
7 years ago
0
Implement asinh magnitudes in `AccuracyBenchmark`
#618
gostevehoward
closed
7 years ago
0
Bring back `error_sds` in AccuracyBenchmark output
#617
gostevehoward
closed
7 years ago
0
pixel counters + don't precompute sky
#616
jeff-regier
closed
7 years ago
0
Missed optimization opportunity in SLP Vectorizer
#615
Keno
closed
7 years ago
2
enable pre-allocated pool of elbo_vars
#614
jrevels
closed
7 years ago
1
Performance improvements
#613
Keno
closed
7 years ago
0
refactor calculate_G_s
#612
jeff-regier
closed
7 years ago
0
steps towards interpolating the sky during optimization rather than during image loading; some refactoring and a unit test to help us call `fill!` less
#611
jeff-regier
closed
7 years ago
0
set the initial trust region radius to 1.0, not 10.
#610
jeff-regier
closed
7 years ago
0
reallocate elbo_vars one per call to maximize\!, not one per evaluation of the objective function
#609
jeff-regier
closed
7 years ago
0
fix Jacobian bounds at compile-time for ConstraintTransform derivatives
#608
jrevels
closed
7 years ago
4
restores the venerable quarter-degree benchmark
#607
jeff-regier
closed
7 years ago
1
recompute `epsilon` as needed, rather than interpolating `epsilon_mat` for the whole sky
#606
jeff-regier
closed
7 years ago
0
don't deepcopy elbo
#605
jeff-regier
closed
7 years ago
0
revising `ElboMaximize` to use a thread-indexed pool of `ElboIntermediateVariables{Float64}` instances
#604
jeff-regier
closed
7 years ago
1
remove allocations from elbo_likelihood
#603
jeff-regier
closed
7 years ago
2
Remove old Stripe82 benchmark code
#602
gostevehoward
closed
7 years ago
2
Write a small set of unit tests for AccuracyBenchmark.jl
#601
gostevehoward
closed
7 years ago
0
split ea into ea, vp, and elbo_vars
#600
jeff-regier
closed
7 years ago
0
porting some improvements from the cg2 branch
#599
jeff-regier
closed
7 years ago
1
removed fft and eda code
#598
jeff-regier
closed
7 years ago
0
instrument code to count the # of (active-source, pixel) pairs and the number (inactive-source, pixel) pairs
#597
jeff-regier
closed
7 years ago
0
Two fixes
#596
Keno
closed
7 years ago
0
added a new benchmark
#595
jeff-regier
closed
7 years ago
0
bump ReverseDiff version
#594
jrevels
closed
7 years ago
3
fix redundant ReverseDiff.compile calls from #587
#593
jrevels
closed
7 years ago
1
minor changes to support running on RCFs outside of Stripe82
#592
gostevehoward
closed
7 years ago
0
Previous
Next