STAN RELEASE NOTES
======================================================================

HOW TO (developer notes on how to release, last column is ordering)
----------------------------------------------------------------------
 1.  Stan C++ M.m.p frozen        (stan/src)                     
 2.  Stan manual M.m.p frozen     (stan/docs/stan-reference/)
 3.  RStan M.m.p frozen           (stan/rstan)                   > 1
 4.  Stan web pages M.m.p frozen  (stan/web)
 5.  tag release vM.m.p in git, push to Google Code              > 1--4
 6.  Build Stan tarball and upload to Google Code->Download      > 5
 7.  Build manual pdf & upload to Google Code->Download          > 5
 8.  Build Rstan & upload to Google Code->Download, Google URL   > 5
 9.  Upload web pages to mc-stan.org                             > 1--8
10.  Announce                                                    > 9


v1.1.0
======================================================================
-- Backward Compatibility Issue
   * Categorical distribution recoded to match documentation;  it
     now has support {1,...,K} rather than {0,...,K-1}.  
   * (RStan) change default value of permuted flag from FALSE to TRUE for
     Stan fit S4 extract() method
-- New Features
   * Conditional (if-then-else) statements
   * While statements
-- New Functions
   * generalized multiply_lower_tri_self_transpose() to non-square
     matrices
   * special functions: log_inv_logit(), log1m_inv_logit()
   * matrix special functions: cumulative_sum()
   * probability functions: poisson_log_log() for log-rate 
     parameterized Poisson
   * matrix functions: block(), diag_pre_multiply(), diag_post_multiply()
   * comparison operators (<, >, <=, >=, ==, !=)
   * boolean operators (!, ||, &&)
   * allow +/- inf values in variable declaration constraints
-- RStan Improvements
   * get_posterior_mean() method for Stan fit objects
   * replaced RcppEigen dependency with include of Eigen source
   * added read_stan_csv() to create Stan fit object from CSV files of
     the form written to disk by the command-line version of Stan
   * as.data.frame() S3 method for Stan fit objects
-- Bug Fixes
   * fixed bug in NUTS diagonal resulting in too small step sizes
   * fixed bug introduced in 1.0.3 that hid line and column number
     bug reporting
   * added checks that data dimensions match as well as sizes
   * removed non-symmetric versions of eigenvalues() and eigenvectors()
   * testing identifiers are not reserved words in C++/Stan
   * trapping/reporting locations of errors in data and init reads
   * improvements in dump data format reader for more R compatibility
     and more generality
   * fix bug in bernoulli logit distro tail density
-- Code Improvements
   * templated out matrix libs to reduce code duplication
   * vectorized auto-dif for tcrossprod() and crossprod()
   * optimizations in Wishart
   * vectorization with efficiency improvements in probability distributions
-- Libraries Updated
   * Eigen version 3.1.1 replaced with version 3.1.2
   * Boost version 1.51.0 replaced with version 1.52.0
-- Manual Improvements
   * New chapter on univariate and multivariate variable transforms
   * Many consistency improvements and typo corrections
   * Information on running command line in parallel from shell



v 1.0.3 Release Notes (25 Oct 2012)
======================================================================
Stan Bug Fixes:
 -- error output in stanc now directed to standard error
 -- catching all exceptions thrown in stanc to prevent core dumps
 -- parser now accepts variable names starting with "int" and "real"
    in first assignment after declaration
 -- now catching unknown truncation cdfs and vectorizations in stanc
 -- (hopefully) less scary message for domain error rejects
 -- removed use of assert(); replaced with exceptions where
    appropriate
 -- fixed boundary condition on reporting max init retry failures
 -- fixed initialization errors in tests that failed on some platforms
 -- improved error checking for matrix size args to prevent segfaults
 -- fixed boundary condition on bounded transform to allow
    beta(0.8,0.8) samples near 0 or 1
Stan New Features:
 -- mdivide_left_tri_low, mdivide_right_tri_low functions
 -- all univariate continuous densities vectorized
 -- better error checking for transforms 
 -- new error return type in stanc for invalid arguments
 -- lower and upper bounds for vectors, row vectors, and matrices
 -- command parameter refresh <= 0 turns off iteration reporting
 -- int divisions produce warnings in stanc
 -- errors caused by invalid transformed parameters include var name
 -- various efficiency improvements and error handling for boundary
    conditions and special cases of probability functions
 -- simplified matrix code with heavier use of templating
Stan Manual Additions:
 -- example of defining new distribution with parameters
 -- diff report vs. previous versions
 -- updated for range constraints on vector, row vector, matrix
 -- updated Eigen licensing information (now MPL, was LGPL)
 -- fixed alpha ordering issue in function index due to bolding
RStan Bug Fixes:
 -- bug of setting parameter epsilon_pm wrongly for NUTS 
 -- fixed seed reset so results match Stan
RStan New Features:
 -- return empty stanfit object upon error such as incorrect data
    input, wrong argument specification, etc. Previously, it will
    stop. The empty stanfit includes the compiled model, so it can be
    reused.
 -- add function stan_demo() to run examples included in Stan. 
 -- add function sflist2stanfit() to merge a list of stanfit objects 
    into one, to facilitate user-level parallelism. 
 -- add S3 method as.matrix, as.array, is.array, dim, 
    dimnames for stanfit objects. 
 -- specify seed using string of digits in addition to numbers  
 -- turn off refresh if value <= 0


V 1.0.2 Release Notes (28 Sep 2012)
----------------------------------------------------------------------
Bug Fixes:
  -- check ranges for LHS of assignment to prevent seg faults
  -- added matrix size and shape tests to avoid seg faults
  -- changed matrix to throw domain_error rather than illegal_argument
  -- removed template variable names from distribution error msgs
  -- indexing fixed for arrays of matrix/vector/row_vector
  -- fixed sd() and variance() to return 0.0 for sequences of size 1
  -- fixed void return type in auto_covariance
  -- patch truncation to return -inf for variates out of range
  -- upgraded BUGS ring model to use constraints plus tan2()
New Features:
  -- print statements
  -- multiply_lower_tri_self_transpose function
  -- vectorized several probability functions (see the manual)
Manual Additions:
  -- programming guide: IRT models 
  -- new appendix: style guide
  -- new appendix: migrating from Stan
  -- new index: function and distribution index


v1.0.1
----------------------------------------------------------------------
Patch release.


v1.0.0
----------------------------------------------------------------------
First release.