search

stephenslab / ebnm

R package to fit Empirical Bayes Normal Means model.
https://stephenslab.github.io/ebnm/
12 stars 9 forks source link

Scale x and s to better handle with very large and small data #44

Closed willwerscheid closed 3 years ago

willwerscheid commented 3 years ago

Closes #38, closes #40. ebnm-paper tests suggest that 5 is a good max radius for trust and max step size for nlm. 10 is still too large, but 1 slows things down a bit too much.

codecov-io commented 3 years ago

Codecov Report

Merging #44 (3118818) into master (d066c13) will increase coverage by 0.13%. The diff coverage is 100.00%.

Impacted file tree graph

@@            Coverage Diff             @@
##           master      #44      +/-   ##
==========================================
+ Coverage   93.16%   93.29%   +0.13%     
==========================================
  Files          14       14              
  Lines        1360     1388      +28     
==========================================
+ Hits         1267     1295      +28     
  Misses         93       93
Impacted Files Coverage Δ
R/ebnm.R 93.83% <100.00%> (+0.11%) :arrow_up:
R/opt_control_defaults.R 100.00% <100.00%> (ø)
R/point_exponential.R 95.38% <100.00%> (+0.12%) :arrow_up:
R/point_laplace.R 97.78% <100.00%> (+0.05%) :arrow_up:
R/point_normal.R 94.25% <100.00%> (+0.14%) :arrow_up:
R/workhorse_normal_mix.R 88.88% <100.00%> (ø)
R/workhorse_parametric.R 99.37% <100.00%> (+0.03%) :arrow_up:

Continue to review full report at Codecov.

Legend - Click here to learn more Δ = absolute <relative> (impact), ø = not affected, ? = missing data Powered by Codecov. Last update 33e1a76...3118818. Read the comment docs.