Code

Multiple improvements to pymcmodels#83

Code

Merged

jchodera merged 24 commits intomasterchoderalab/assaytools:masterfrom improvementschoderalab/assaytools:improvementsCopy head branch name to clipboardApr 22, 2017

+354-172

Commits

Commits on Apr 17, 2017

• Add support for specified F_PL and its uncertainty dF_PL

  jchodera

  committed
  f5b5f51

  Copy full SHA for f5b5f51

Commits on Apr 18, 2017

• Remove 'normal' concentration priors; use LogNormalWrapper to simplify lognormal priors.

  jchodera

  committed
  0bdb607

  Copy full SHA for 0bdb607

• Clean up LogNormalWrapper

  jchodera

  committed
  d17a381

  Copy full SHA for d17a381

• Convert quantum yield priors to lognormal by default.

  jchodera

  committed
  4814199

  Copy full SHA for 4814199

• Convert fluorescence and absorbance errors to lognormal

  jchodera

  committed
  0d3e197

  Copy full SHA for 0d3e197

• Fix py3 issues in quickmodel

  jchodera

  committed
  2bc6031

  Copy full SHA for 2bc6031

Commits on Apr 19, 2017

• Fix python 3 incompatibility in platereader.py

  jchodera

  committed
  8c89cd6

  Copy full SHA for 8c89cd6

• Have quickmodel print exceptions it runs into

  jchodera

  committed
  c7697ad

  Copy full SHA for c7697ad

• Adjust Metropolis step methods for new LogNormal priors; some py3 fixes

  jchodera

  committed
  99a0839

  Copy full SHA for 99a0839

• Start [L]=0 implementation

  jchodera

  committed
  85e15da

  Copy full SHA for 85e15da

• Merge branch 'improvements' of github.com:choderalab/assaytools into improvements

  jchodera

  committed
  cf69d59

  Copy full SHA for cf69d59

• Add quickmodel to continuous integration tests.

  jchodera

  committed
  3e99202

  Copy full SHA for 3e99202

• Allow ligand or protein concentrations to be zero.

  jchodera

  committed
  fd9724e

  Copy full SHA for fd9724e

• Add command-line control over number of samples and thinning for quickmodel

  jchodera

  committed
  b3a6dcc

  Copy full SHA for b3a6dcc

• Restore close to previous defaults for quickmodel

  jchodera

  committed
  e1f738a

  Copy full SHA for e1f738a

• Add pymbar to requirements

  jchodera

  committed
  b1026a4

  Copy full SHA for b1026a4

• Whoops! Previous behavior was actually 10000 samples, not 1000

  jchodera

  committed
  f26e8c0

  Copy full SHA for f26e8c0

• quickmodel now plots equilibrated traces in a different shade; fix bug in command-line argument parsing

  jchodera

  committed
  ef095fe

  Copy full SHA for ef095fe

• Fixed bug where same ligand concentration was being used for both +/- protein rows

  jchodera

  committed
  f41dc66

  Copy full SHA for f41dc66

• Decouple +/- protein buffer background priors

  jchodera

  committed
  20a3e81

  Copy full SHA for 20a3e81

• Restored Metropolis step methods with better step size guesses

  jchodera

  committed
  dbec19f

  Copy full SHA for dbec19f

Commits on Apr 20, 2017

• Add AdaptiveMetropolis for correlated DeltaG and log_F_PL moves

  jchodera

  committed
  b02d9ac

  Copy full SHA for b02d9ac

• Added tune_throughout=True

  jchodera

  committed
  d841d8f

  Copy full SHA for d841d8f

• Allow Metropolis methods to tune throughout

  jchodera

  committed
  fcde1cf

  Copy full SHA for fcde1cf