The posterior R package is intended to provide useful tools for both users and developers of packages for fitting Bayesian models or working with output from Bayesian models. The primary goals of the package are to:
If you are new to posterior we recommend starting with these vignettes:
You can install the latest official release version via
install.packages("posterior")
or build the developmental version directly from GitHub via
# install.packages("remotes")
remotes::install_github("stan-dev/posterior")
Here we offer a few examples of using the package. For a more detailed overview see the vignette The posterior R package.
library("posterior")
#> This is posterior version 1.4.1.9001
#>
#> Attaching package: 'posterior'
#> The following objects are masked from 'package:stats':
#>
#> mad, sd, var
#> The following objects are masked from 'package:base':
#>
#> %in%, match
To demonstrate how to work with the posterior package, we will use
example posterior draws obtained from the eight schools hierarchical
meta-analysis model described in Gelman et al. (2013). Essentially, we
have an estimate per school (theta[1]
through theta[8]
) as well as
an overall mean (mu
) and standard deviation across schools (tau
).
eight_schools_array <- example_draws("eight_schools")
print(eight_schools_array, max_variables = 3)
#> # A draws_array: 100 iterations, 4 chains, and 10 variables
#> , , variable = mu
#>
#> chain
#> iteration 1 2 3 4
#> 1 2.0 3.0 1.79 6.5
#> 2 1.5 8.2 5.99 9.1
#> 3 5.8 -1.2 2.56 0.2
#> 4 6.8 10.9 2.79 3.7
#> 5 1.8 9.8 -0.03 5.5
#>
#> , , variable = tau
#>
#> chain
#> iteration 1 2 3 4
#> 1 2.8 2.80 8.7 3.8
#> 2 7.0 2.76 2.9 6.8
#> 3 9.7 0.57 8.4 5.3
#> 4 4.8 2.45 4.4 1.6
#> 5 2.8 2.80 11.0 3.0
#>
#> , , variable = theta[1]
#>
#> chain
#> iteration 1 2 3 4
#> 1 3.96 6.26 13.3 5.78
#> 2 0.12 9.32 6.3 2.09
#> 3 21.25 -0.97 10.6 15.72
#> 4 14.70 12.45 5.4 2.69
#> 5 5.96 9.75 8.2 -0.91
#>
#> # ... with 95 more iterations, and 7 more variables
The draws for this example come as a draws_array
object, that is, an
array with dimensions iterations x chains x variables. We can easily
transform it to another format, for instance, a data frame with
additional meta information.
eight_schools_df <- as_draws_df(eight_schools_array)
print(eight_schools_df)
#> # A draws_df: 100 iterations, 4 chains, and 10 variables
#> mu tau theta[1] theta[2] theta[3] theta[4] theta[5] theta[6]
#> 1 2.01 2.8 3.96 0.271 -0.74 2.1 0.923 1.7
#> 2 1.46 7.0 0.12 -0.069 0.95 7.3 -0.062 11.3
#> 3 5.81 9.7 21.25 14.931 1.83 1.4 0.531 7.2
#> 4 6.85 4.8 14.70 8.586 2.67 4.4 4.758 8.1
#> 5 1.81 2.8 5.96 1.156 3.11 2.0 0.769 4.7
#> 6 3.84 4.1 5.76 9.909 -1.00 5.3 5.889 -1.7
#> 7 5.47 4.0 4.03 4.151 10.15 6.6 3.741 -2.2
#> 8 1.20 1.5 -0.28 1.846 0.47 4.3 1.467 3.3
#> 9 0.15 3.9 1.81 0.661 0.86 4.5 -1.025 1.1
#> 10 7.17 1.8 6.08 8.102 7.68 5.6 7.106 8.5
#> # ... with 390 more draws, and 2 more variables
#> # ... hidden reserved variables {'.chain', '.iteration', '.draw'}
Different formats are preferable in different situations and hence
posterior supports multiple formats and easy conversion between them.
For more details on the available formats see help("draws")
. All of
the formats are essentially base R object classes and can be used as
such. For example, a draws_matrix
object is just a matrix
with a
little more consistency and additional methods.
Computing summaries of posterior or prior draws and convergence
diagnostics for posterior draws is one of the most common tasks when
working with Bayesian models fit using Markov Chain Monte Carlo (MCMC)
methods. The posterior package provides a flexible interface for
this purpose via summarise_draws()
:
# summarise_draws or summarize_draws
summarise_draws(eight_schools_df)
#> # A tibble: 10 × 10
#> variable mean median sd mad q5 q95 rhat ess_bulk ess_tail
#> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 mu 4.18 4.16 3.40 3.57 -0.854 9.39 1.02 558. 322.
#> 2 tau 4.16 3.07 3.58 2.89 0.309 11.0 1.01 246. 202.
#> 3 theta[1] 6.75 5.97 6.30 4.87 -1.23 18.9 1.01 400. 254.
#> 4 theta[2] 5.25 5.13 4.63 4.25 -1.97 12.5 1.02 564. 372.
#> 5 theta[3] 3.04 3.99 6.80 4.94 -10.3 11.9 1.01 312. 205.
#> 6 theta[4] 4.86 4.99 4.92 4.51 -3.57 12.2 1.02 695. 252.
#> 7 theta[5] 3.22 3.72 5.08 4.38 -5.93 10.8 1.01 523. 306.
#> 8 theta[6] 3.99 4.14 5.16 4.81 -4.32 11.5 1.02 548. 205.
#> 9 theta[7] 6.50 5.90 5.26 4.54 -1.19 15.4 1.00 434. 308.
#> 10 theta[8] 4.57 4.64 5.25 4.89 -3.79 12.2 1.02 355. 146.
Basically, we get a data frame with one row per variable and one column
per summary statistic or convergence diagnostic. The summaries rhat
,
ess_bulk
, and ess_tail
are described in Vehtari et al. (2020). We
can choose which summaries to compute by passing additional arguments,
either functions or names of functions. For instance, if we only wanted
the mean and its corresponding Monte Carlo Standard Error (MCSE) we
would use:
summarise_draws(eight_schools_df, "mean", "mcse_mean")
#> # A tibble: 10 × 3
#> variable mean mcse_mean
#> <chr> <dbl> <dbl>
#> 1 mu 4.18 0.150
#> 2 tau 4.16 0.213
#> 3 theta[1] 6.75 0.319
#> 4 theta[2] 5.25 0.202
#> 5 theta[3] 3.04 0.447
#> 6 theta[4] 4.86 0.189
#> 7 theta[5] 3.22 0.232
#> 8 theta[6] 3.99 0.222
#> 9 theta[7] 6.50 0.250
#> 10 theta[8] 4.57 0.273
For a function to work with summarise_draws
, it needs to take a vector
or matrix of numeric values and returns a single numeric value or a
named vector of numeric values.
Another common task when working with posterior (or prior) draws, is
subsetting according to various aspects of the draws (iterations,
chains, or variables). posterior provides a convenient interface for
this purpose via the subset_draws()
method. For example, here is the
code to extract the first five iterations of the first two chains of the
variable mu
:
subset_draws(eight_schools_df, variable = "mu", chain = 1:2, iteration = 1:5)
#> # A draws_df: 5 iterations, 2 chains, and 1 variables
#> mu
#> 1 2.0
#> 2 1.5
#> 3 5.8
#> 4 6.8
#> 5 1.8
#> 6 3.0
#> 7 8.2
#> 8 -1.2
#> 9 10.9
#> 10 9.8
#> # ... hidden reserved variables {'.chain', '.iteration', '.draw'}
The same call to subset_draws()
can be used regardless of whether the
object is a draws_df
, draws_array
, draws_list
, etc.
The magic of having obtained draws from the joint posterior (or prior)
distribution of a set of variables is that these draws can also be used
to obtain draws from any other variable that is a function of the
original variables. That is, if are interested in the posterior
distribution of, say, phi = (mu + tau)^2
all we have to do is to
perform the transformation for each of the individual draws to obtain
draws from the posterior distribution of the transformed variable. This
procedure is automated in the mutate_variables
method:
x <- mutate_variables(eight_schools_df, phi = (mu + tau)^2)
x <- subset_draws(x, c("mu", "tau", "phi"))
print(x)
#> # A draws_df: 100 iterations, 4 chains, and 3 variables
#> mu tau phi
#> 1 2.01 2.8 22.8
#> 2 1.46 7.0 71.2
#> 3 5.81 9.7 240.0
#> 4 6.85 4.8 135.4
#> 5 1.81 2.8 21.7
#> 6 3.84 4.1 62.8
#> 7 5.47 4.0 88.8
#> 8 1.20 1.5 7.1
#> 9 0.15 3.9 16.6
#> 10 7.17 1.8 79.9
#> # ... with 390 more draws
#> # ... hidden reserved variables {'.chain', '.iteration', '.draw'}
When we do the math ourselves, we see that indeed for each draw, phi
is equal to (mu + tau)^2
(up to rounding two 2 digits for the purpose
of printing).
We may also easily rename variables, or even entire vectors of variables
via rename_variables
, for example:
x <- rename_variables(eight_schools_df, mean = mu, alpha = theta)
variables(x)
#> [1] "mean" "tau" "alpha[1]" "alpha[2]" "alpha[3]" "alpha[4]" "alpha[5]"
#> [8] "alpha[6]" "alpha[7]" "alpha[8]"
As with all posterior methods, mutate_variables
and
rename_variables
can be used with all draws formats.
Suppose we have multiple draws objects that we want to bind together:
x1 <- draws_matrix(alpha = rnorm(5), beta = 1)
x2 <- draws_matrix(alpha = rnorm(5), beta = 2)
x3 <- draws_matrix(theta = rexp(5))
Then, we can use the bind_draws
method to bind them along different
dimensions. For example, we can bind x1
and x3
together along the
'variable'
dimension:
x4 <- bind_draws(x1, x3, along = "variable")
print(x4)
#> # A draws_matrix: 5 iterations, 1 chains, and 3 variables
#> variable
#> draw alpha beta theta
#> 1 0.804 1 1.296
#> 2 0.078 1 0.983
#> 3 1.687 1 0.099
#> 4 -1.906 1 0.094
#> 5 -0.072 1 2.189
Or, we can bind x1
and x2
together along the 'draw'
dimension:
x5 <- bind_draws(x1, x2, along = "draw")
print(x5)
#> # A draws_matrix: 10 iterations, 1 chains, and 2 variables
#> variable
#> draw alpha beta
#> 1 0.804 1
#> 2 0.078 1
#> 3 1.687 1
#> 4 -1.906 1
#> 5 -0.072 1
#> 6 0.654 2
#> 7 0.675 2
#> 8 0.104 2
#> 9 -0.403 2
#> 10 2.274 2
As with all posterior methods, bind_draws
can be used with all
draws formats.
The eight_schools
example already comes in a format natively supported
by posterior but we could of course also import the draws from other
sources, for example, from common base R objects:
x <- matrix(rnorm(50), nrow = 10, ncol = 5)
colnames(x) <- paste0("V", 1:5)
x <- as_draws_matrix(x)
print(x)
#> # A draws_matrix: 10 iterations, 1 chains, and 5 variables
#> variable
#> draw V1 V2 V3 V4 V5
#> 1 -0.51 0.49 0.49 -0.39 -2.828
#> 2 1.02 -0.41 -1.19 -0.54 0.711
#> 3 -2.83 1.28 -1.04 0.30 1.032
#> 4 -1.61 0.39 1.08 1.22 -1.564
#> 5 0.47 -0.50 0.40 0.13 0.202
#> 6 -1.51 0.37 0.79 -1.55 -0.346
#> 7 1.83 -2.90 -0.28 0.49 1.206
#> 8 -1.09 0.86 -1.52 -0.32 0.098
#> 9 0.21 -1.79 0.25 0.16 -1.371
#> 10 1.44 -0.92 -0.52 0.22 0.889
summarise_draws(x, "mean", "sd", "median", "mad")
#> # A tibble: 5 × 5
#> variable mean sd median mad
#> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 V1 -0.257 1.51 -0.147 1.88
#> 2 V2 -0.311 1.28 -0.0179 1.03
#> 3 V3 -0.154 0.893 -0.0154 0.972
#> 4 V4 -0.0296 0.733 0.144 0.598
#> 5 V5 -0.197 1.33 0.150 1.20
Instead of as_draws_matrix()
we also could have just used
as_draws()
, which attempts to find the closest available format to the
input object. In this case this would result in a draws_matrix
object
either way.
The above matrix example contained only one chain. Multi-chain draws could be stored in base R 3-D array object, which can also be converted to a draws object:
x <- array(data=rnorm(200), dim=c(10, 2, 5))
x <- as_draws_matrix(x)
variables(x) <- paste0("V", 1:5)
print(x)
#> # A draws_matrix: 10 iterations, 2 chains, and 5 variables
#> variable
#> draw V1 V2 V3 V4 V5
#> 1 -1.24 1.39 0.946 0.2391 -0.68
#> 2 0.46 -0.14 0.852 -0.0092 0.67
#> 3 0.20 -1.35 -0.388 -1.2112 -0.29
#> 4 -0.77 -0.61 -0.649 -1.4109 -1.10
#> 5 -0.45 -0.12 -1.095 -1.3789 1.28
#> 6 -1.01 1.40 -0.072 0.6276 -0.57
#> 7 0.28 0.58 -0.248 0.1293 0.39
#> 8 -0.22 -0.83 2.084 0.0950 0.76
#> 9 -0.39 1.99 0.914 0.9035 -0.65
#> 10 -1.87 0.66 -0.505 -1.4585 0.42
#> # ... with 10 more draws
The coda and rjags packages use mcmc
and mcmc.list
objects
which can also be converted to draws objects:
data(line, package = "coda")
line <- as_draws_df(line)
print(line)
#> # A draws_df: 200 iterations, 2 chains, and 3 variables
#> alpha beta sigma
#> 1 7.2 -1.57 11.23
#> 2 3.0 1.50 4.89
#> 3 3.7 0.63 1.40
#> 4 3.3 1.18 0.66
#> 5 3.7 0.49 1.36
#> 6 3.6 0.21 1.04
#> 7 2.7 0.88 1.29
#> 8 3.0 1.09 0.46
#> 9 3.5 1.07 0.63
#> 10 2.1 1.48 0.91
#> # ... with 390 more draws
#> # ... hidden reserved variables {'.chain', '.iteration', '.draw'}
We welcome contributions! The posterior package is under active development. If you find bugs or have ideas for new features (for us or yourself to implement) please open an issue on GitHub (https://github.com/stan-dev/posterior/issues).
Developing and maintaining open source software is an important yet often underappreciated contribution to scientific progress. Thus, whenever you are using open source software (or software in general), please make sure to cite it appropriately so that developers get credit for their work.
When using posterior, please cite it as follows:
When using the MCMC convergence diagnostics rhat
, ess_bulk
, or
ess_tail
, please also cite
The same information can be obtained by running citation("posterior")
.
Gelman A., Carlin J. B., Stern H. S., David B. Dunson D. B., Aki Vehtari A., & Rubin D. B. (2013). Bayesian Data Analysis, Third Edition. Chapman and Hall/CRC.
Vehtari A., Gelman A., Simpson D., Carpenter B., & Bürkner P. C. (2021). Rank-normalization, folding, and localization: An improved Rhat for assessing convergence of MCMC (with discussion). Bayesian Analysis. 16(2), 667–718. doi.org/10.1214/20-BA1221
The posterior package is licensed under the following licenses: