adding single-sample normalization (noob with recriprocation) for 27k/450k/EPIC merging

hansenlab / minfi

Devel repository for minfi

58 stars 70 forks source link

adding single-sample normalization (noob with recriprocation) for 27k/450k/EPIC merging #65

Closed ttriche closed 8 years ago

ttriche commented 8 years ago

In general, the differences are tiny, but this avoids screwing up p-values and allows individual pairs of IDATs to be incrementally normalized & added to an existing corpus. This can be handy when adding a few dozen or a few hundred new samples to an existing dataset of a few thousand. It also means the entire operation can be performed distributed and/or in parallel without batch dependencies.

codecov-io commented 8 years ago

Current coverage is 30.73%

Merging #65 into master will increase coverage by 0.11%

@@             master        #65   diff @@
==========================================
  Files            30         30          
  Lines          3165       3179    +14   
  Methods           0          0          
  Messages          0          0          
  Branches          0          0          
==========================================
+ Hits            969        977     +8   
- Misses         2196       2202     +6   
  Partials          0          0

Powered by Codecov. Last updated by 85e070d...b29ac4a