NEWS.md

In eheinzen/gibbs.utils: Utilities for Gibbs Sampling

gibbs.utils v5.0.2

• Fixed a bug in bss() when x= wasn't spanning all its possible values.

gibbs.utils v5.0.1

• Allowed negative infinity means in sample_pois_reg() in some circumstances.

gibbs.utils v5.0.0

• Removed times_flat_ar1(), which wasn't well-documented and no longer all that useful to me.

• Added bss() and friends.

gibbs.utils v4.9.9

• Fixed a bug in the conjugacy in conj_gamma_b().

gibbs.utils v4.9.8

• Added the argument b= to dmvnorm_diff() for the canonical representation.

gibbs.utils v4.9.7

• Enforce spam >= 2.9-2 because of a critical bug fix related to the element-wise multiplication of a full spam object. This led to incorrect densities. Note: if you can't find a compatible version on CRAN, try the source GitLab, which has tarballs of the packages.

• Added the diag= argument to conj_matnorm_mu() and conj_matlm_beta() output when params.only=TRUE.

• diag=TRUE now works for gs_matlm_beta(). It's slightly slower when the dimensions are small, but MUCH faster when the dimensions get large.

gibbs.utils v4.9.6

• Fixed a bug in conj_matlm_beta() and conj_matnorm_mu() when diag=TRUE and a dimension is 1.

gibbs.utils v4.9.5

• Changed a few more base::diag() to spam::diag() (I missed these last time).

gibbs.utils v4.9.4

• Changed a few more base::diag() to spam::diag() in the case diag=TRUE and Q0 or precision is a spam object.

• Changed some internals for sample_multinom_reg() when diag=TRUE. In particular, precision can now be a spam matrix.

gibbs.utils v4.9.3

• Changed base::diag() to spam::diag() in conj_matlm_mu() and conj_matlm_beta(), in the case diag=TRUE and Q0 is a spam object.

gibbs.utils v4.9.2

• Added [dpr]dlnorm() for the discrete log-normal distribution.

gibbs.utils v4.9.1

• Added conj_gamma_b().

gibbs.utils v4.9.0

• Fixed a misspecified glue.

• Changed all the relevant C++ code to include a const declaration, so I don't regret it later.

• Removed MASS as a dependency.

gibbs.utils v4.8.1

• Added the special case mu=NULL to dmatnorm(), dmvnorm(), and dmvlnorm().

• Fixed an issue in the multinomial approximations when the number of rows is 1.

gibbs.utils v4.8.0

• Removed some spurious code.

• Added rtruncexp(), qtruncexp(), ptruncexp(), and dtruncexp() for truncated exponential distributions.

• Added "mv truncated exponential" to sample_pois_reg(), sample_binom_reg(), and sample_multinom_reg(), to approximate the log-density with a first-order Taylor approximation (which is an exponential distribution).

• Fixed a bug in the "mv beta" method of sample_binom_reg(), where a + sign was accidentally present in place of a -.

• Fixed the pkgdown changelog by adding "gibbs.utils" to all the headers so it parses correctly.

gibbs.utils v4.7.0

• Added method to sample_multinom_reg() for "mv ind quadratic taylor".

• The width= argument is now squared when applied to the variance for various methods, so that it is consistency on the sd scale.

gibbs.utils v4.6.0

• Sped up sample_pois_reg() and sample_binom_reg() by something like half, and sample_multinom_reg() by maybe 1 percent by improving matrix subscripting in the C++ code.

• Improved the numeric stability of the gamma approximations in sample_pois_reg() and the beta approximations in sample_binom_reg().

• Fixed a bug in the (still experimental) "mv beta" method of sample_binom_reg().

• Added methods to sample_multinom_reg() for "normal", "uniform", and "quadratic taylor".

gibbs.utils v4.5.1

• Sped up sample_multinom_reg() by 7x by reworking some C++ internals.

gibbs.utils v4.5.0

• Added rdirich() to generate random Dirichlet vectors.

gibbs.utils v4.4.0

• Added the times_flat_ar1() function, whose details will undoubtedly come in another release.

• Documented the diag= argument from v4.3.0.

gibbs.utils v4.3.0

• Added the diag= argument to sample_multinom_reg(), for when the precision matrix is diagonal, yielding a speedup (because it can skip calculating the conditional mean).

gibbs.utils v4.2.1

• Sped up conj_matlm_beta() (by about 4x) and conj_matnorm_mu() (very slightly) by only matrix multiplying once when diag=TRUE.

gibbs.utils v4.2.0

• Added support for mu=NULL in conj_matnorm_V() and conj_mvnorm_Q().

gibbs.utils v4.1.0

• Added the zmax= argument to sample_multinom_reg(), for advanced usage.

gibbs.utils v4.0.0

• Overhauled sample_multinom_reg(). Where possible, backwards compatibility was maintained, but results will differ because of random seeds and whatnot.

gibbs.utils v3.0.0

• Added truncate= argument to sample_pois_reg(), to truncate the Poisson likelihood.

gibbs.utils v2.5.0

• Added dmvlnorm(), dmvlnorm_diff(), dnorm_diff(), dlnorm_diff().

• Added an argument byrow= to dmvnorm_diff().

• Critical bug fix for "[mv ]gamma" in sample_pois_reg() and "mv beta" in sample_binom_reg().

gibbs.utils v2.4.0

• Added "mv beta" to sample_binom_reg().

• Added "mv ind quadratic taylor" to sample_binom_reg() and sample_pois_reg().

• Changed the accept_regardless= argument to acceptance=, where one option ("LL only") is to accept ignoring the proposal density.

gibbs.utils v2.3.1

• Some speed improvements for "mv quadratic taylor" and "mv gamma".

gibbs.utils v2.3.0

• sample_pois_reg():

  • Fixed a bug where "gamma" didn't actually run the gamma proposal.

  • Fixed the "gamma" metropolis step in cases when mult != 1.

  • Added a "mv gamma" option.

• Added accept_regardless= argument to sample_pois_reg() and sample_binom_reg()

gibbs.utils v2.2.0

• Removed chol_mvrnorm() and all its references. spam:: functions are faster.

gibbs.utils v2.1.0

• Saved some time by using C++ pass-by-reference.

• Added "gamma" proposal to sample_pois_reg().

gibbs.utils v2.0.0

• ss_pois_reg() and mh_pois_reg() now have a unified interface: sample_pois_reg(), with slightly different naming conventions for extra arguments.

• ss_binom_reg() and mh_binom_reg() now have a unified interface: sample_binom_reg(), with slightly different naming conventions for extra arguments.

• ss_multinom_reg() was renamed to sample_multinom_reg() for consistency, with slightly different naming conventions for extra arguments, and the additional argument method= for future development.

gibbs.utils v1.4.3

• ss_multinom_reg() now allows a 3-dimensional z= argument, for a small speed penalty.

gibbs.utils v1.4.2

• Added dmatnorm_diff(), akin to dmvnorm_diff().

gibbs.utils v1.4.1

• Added dmvnorm_diff() for taking the difference of two (log-) densities, useful in, e.g. Metropolis-Hastings sampling, but faster than taking individual densities twice.

gibbs.utils v1.4.0

• Removed conj_matlm_V(), which is nearly an exact duplicate of conj_matnorm_V().

• Added some shortcut arguments for t(X) %*% U, t(X) %*% U %*% X, and t(ymu) %*% U %*% ymu to conj_matlm_beta() and conj_matnorm_V().

gibbs.utils v1.3.2

• options(spam.structurebased=FALSE).

gibbs.utils v1.3.1

• Replaced as.numeric() with as.vector(), for compatibility with spam.

gibbs.utils v1.3.0

• Replaced gs_diagmatlm_beta() with gs_matlm_beta(), which now passes arguments to conj_matlm_beta() through the dots. This includes the new zero= argument, which can be used to recover the performance of gs_diagmatlm_beta().

gibbs.utils v1.2.1

• I accidentally padded the results from the changes in v1.2.0 to include the structural zeros, when they shouldn't have.

gibbs.utils v1.2.0

• Added the zero= argument to conj_matnorm_mu() and conj_matlm_beta() to indicate structural zeros. As such, we removed conj_diagmatlm_beta(), which was slightly faster with small inputs, and about the same speed with larger ones. (Author's note: implementing a special condition for when zero is a diagonal matrix does remedy the small inputs speed difference, but I decided it wasn't worth maintaining more code for something that's already very fast).

• Added the derivation of the above conjugate case to the vignette.

gibbs.utils v1.1.1

• Sped up conj_diagmatlm_beta() (and hence gs_diagmatlm_beta()) by taking the linear algebra a few steps further.

gibbs.utils v1.1.0

• Added is.gu_chol(), and now allow for passing of newQ.chol= to be taken into account when sampling with spam::rmvnorm.canonical() (by appending the gu_chol class, which doesn't take the Cholesky another time). Notably, this doesn't work for spam objects, because they are S4.

• chol.gu_chol() gained a verbose= argument (mostly for debugging).

• gu_chol() gained a take.chol= argument, to mark existing Cholesky matrices as such without taking chol() again.

• Fixed a bug when params.only=TRUE and the input newQ.chol= is just a regular matrix.

gibbs.utils v1.0.0

• Overhauled the conjugate normal functions, to use spam::rmvnorm.canonical() instead of taking two inverses. This method is slower than what we had before when inputs are really small (when the conjugacy is super fast anyway, so I wasn't worried) but almost 2x faster when inputs are big. Furthermore, we can get more performance when inputs are spam objects already.

• spam was added as a dependency.

gibbs.utils v0.10.9

• Replaced crossprod(x, y) with t(x) %*% y in almost all instances, to be compatible with, e.g., the spam package for sparse matrices.

• Fixed a typo in the vignette

gibbs.utils v0.10.8

• Added arguments for pre-computed determinants to dmvnorm() and dmatnorm().

gibbs.utils v0.10.7

• Applied the same vectorization trick to dmatnorm(), for speed gains of over 1000x for large inputs. The default for U= is also now NULL, indicating the identity matrix.

gibbs.utils v0.10.6

• Removed the use_trace= argument introduced in v0.9.3. When x was short, the trace was faster. When x= was long, the non-trace method was faster. Now, using a vectorization trick with the trace, dmvnorm() is always at least as fast (and usually faster) than either of the previous methods.

gibbs.utils v0.10.5

• Also return V.inv for conjugate Wishart parameters, tau for conjugate normal, and Q for some other normal conjugate cases where newQ.inv= has not already been supplied.

• Changed the defaults (but not the default behavior) for some normal conjugates.

gibbs.utils v0.10.4

• Changed the default for the use_trace= argument introduced in v0.9.3, because in my experience it's faster not to use the matrix trace.

gibbs.utils v0.10.3

• Add the same mu0 = NULL default to gs_diagmatlm_beta().

gibbs.utils v0.10.2

• Speed up (slightly) the conjugacy functions using a default of mu0 = NULL to indicate zero. This also allows you to give infinite precisions when mu0 = NULL = 0 (which would otherwise give Inf * 0 = NaN).

gibbs.utils v0.10.1

• Export mvqt_binom_approx() and mvqt_pois_approx().

gibbs.utils v0.10.0

• Added mh_pois_reg().

• Added proposal="mv quadratic taylor" to mh_binom_reg() to do multivariate Taylor's approximations and MH steps.

gibbs.utils v0.9.3

• Added an option to dmvnorm() to opt out of the matrix trace.

gibbs.utils v0.9.2

• Sped up dmvnorm() by using the matrix trace instead of a for-loop.

gibbs.utils v0.9.1

• Changed the n == k == 0 sampling to C++ for ss_binom_reg() and mh_binom_reg()

• ss_pois_reg() now accepts NA in k=, to indicate that no Poisson draw was observed.

gibbs.utils v0.9.0

• Renamed conj_matlm_sigma() to conj_matlm_V(). It, along with conj_matnorm_V(), now both use conj_mvnorm_Q() in the case that U = NULL (independence among rows).

gibbs.utils v0.8.2

• Added truncnorm as a package dependency.

• Added gs_diagmatlm_beta() to Gibbs-sample truncated normal diagonal matrix-normal regression means.

• cond_mvnorm() gained optional truncation parameters

• Updated the vignettes to improve the custom LaTeX code that wasn't previously handled by HTML output.

• Improved printing of "gibbs_utils_params" objects.

gibbs.utils v0.8.1

• conj_matnorm_mu() and conj_matlm_beta() gained the diag= argument, for 100x speed boost when both V= and Q0= are diagonal.

• When params.only=TRUE, the mu vector returned now drops any dimensions in conj_matnorm_mu(), conj_matlm_beta(), conj_mvnorm_mu(), conj_diagmatlm_beta(), and conj_lm_beta()

• When n==1, chol_mvrnorm() now drops dimensions, to match what MASS::mvrnorm() does.

• Improved printing of "gibbs_utils_params" objects.

gibbs.utils v0.8.0

• ss_multinom_reg() now also uses normal draws in the case where z == 0 or n == 0. This is not done multivariately, but rather univariately.

• ss_multinom_reg() now accepts numeric values for the ref= argument, to allow inner "columns" to be the reference.

gibbs.utils v0.7.0

• Sped up the multivariate binomial, multinomial, and Poisson likelihood evaluations (only completing the square once instead of each time), so that

  • ss_pois_reg() is as much as 6.5x faster

  • ss_multinom_reg() is as much as 1.6x faster

  • ss_binom_reg() is as much as 4.5x faster

  • mh_binom_reg() is as much as 2x faster.

As such, there are no longer C++ functions to evaluate the multivariate log-likelihoods (nor the slice sample functions to do the same); instead, the conditional means are computed within cond_mv_mean(), and the results passed to the univariate likelihoods.

• Overhauled mh_binom_reg(). It now takes a proposal= argument, to determine how proposals are made. The "normal" method is the default for backwards-compatibility, but the "quadratic taylor" gives better results and is about the same speed. The C++ internals have also been almost completely rewritten.

• Updated mh_binom_reg() and ss_binom_reg() to take normal draws in the multivariate case when only one n= is zero instead of only doing it when all the n='s are zero.

• mh_binom_reg() now enforces k <= n

• Fixed one critical bug in mh_binom_reg() in which the log-likelihood evaluations were using k= instead of n=.

gibbs.utils v0.6.0

• ss_binom_reg() now performs conjugate normal updates when n == 0 (or rowSums(n) == 0 in the multivariate case).

• ss_binom_reg() now enforces k <= n

• Fixed a bug in chol_mvrnorm().

gibbs.utils v0.5.1

• Small performance improvements for conj_mvnorm().

gibbs.utils v0.5.0

• Added conj_binom_p() for the beta conjugacy for a binomial probability.

• Fixed a bug in ss_multinom_reg() when dimensions were length 1.

gibbs.utils v0.4.8

• A bug fix for slice sampling multinomial.

gibbs.utils v0.4.7

• Added some arguments for speed and efficiency, and therefore removed some parameters if params.only = TRUE.

gibbs.utils v0.4.6

• Updated vignettes.

• Added conj_diagmatlm_beta().

• Added tests for conjugacy.

gibbs.utils v0.4.5

• Added ref= argument to ss_multinom_reg().

gibbs.utils v0.4.4

• Sped up multinomial log-likelihood by avoiding exponentiation altogether for structural zeros.

• Changed the acceptable dimensions of mean= and precision= in ss_multinom_reg()

gibbs.utils v0.4.3

• Sped up sampling (by maybe 15%?) by avoiding repeated subsetting of k and n.

• Added ss_multinom_reg().

gibbs.utils v0.4.2

• Sped up binomial log-likelihood (by a non-trivial amount) by avoiding an exponentiation.

gibbs.utils v0.4.1

• Added a pkgdown site (and README.md).

• Added the params.only= argument to conjugacy functions.

• Added impute_conj_mvnorm().

gibbs.utils v0.4.0

• Include an option to use (the new function) chol_mvrnorm() to draw multivariate normal samples instead of MASS::mvrnorm() (which uses eigen decomposition instead of Cholesky).

gibbs.utils v0.3.2

• Defer to stats::rWishart() instead of rwish(), since the former is faster.

• Added mh_binom_reg() (and proceeded to fix one critical bug in 61de4f70).

• Sped up multivariate log-likelihood.

• Fixed one test.

gibbs.utils v0.3.1

• Updated ss_pois_reg() and ss_binom_reg() to allow for matrix (multivariate) inputs.

• Added tests.

gibbs.utils v0.3.0

• Updated ss_binom_reg() to take in and output logit-p instead of p.

gibbs.utils v0.2.0

• Improved matrix normal conjugacies by using a Kronecker product trick.

gibbs.utils v0.1.2

• Fixed a bug in defaults for conj_mvnorm_mu(), where the length was accidentally the number of observations instead of variables.

• Improved documentation.

• Improved checking for matrix arguments.

gibbs.utils v0.1.1

• Export logit() and expit()

• Rename my_inv() to chol_inv().

gibbs.utils v0.1.0

Initial version of package

eheinzen/gibbs.utils documentation built on Sept. 27, 2024, 9:03 p.m.