Description Usage Arguments Value Author(s) See Also Examples
View source: R/calc_hpd_interval.R
Calculate the Highest Probability Density of an MCMC trace that has its burn-in removed
1 | calc_hpd_interval(trace, proportion = 0.95)
|
trace |
a numeric vector of parameter estimates obtained from an MCMC run. Must have its burn-in removed |
proportion |
the proportion of numbers within the interval. For example, use 0.95 for a 95 percentage interval |
a numeric vector, with at index 1 the lower boundary of the interval, and at index 2 the upper boundary of the interval
The original Java version of the algorithm was from J. Heled, ported to R and adapted by Richèl J.C. Bilderbeek
The function remove_burn_in
removes
a burn-in.
The Java code that inspired this function can be found here:
https://github.com/beast-dev/beast-mcmc/blob/98705c59db65e4f406a420bbade949aeecfe05d0/src/dr/stats/DiscreteStatistics.java#L317 # nolint URLs can be long
1 2 3 4 5 6 7 8 9 10 | estimates <- parse_beast_tracelog_file(
get_tracerer_path("beast2_example_output.log")
)
tree_height_trace <- remove_burn_in(
estimates$TreeHeight,
burn_in_fraction = 0.1
)
# Values will be 0.453 and 1.816
calc_hpd_interval(tree_height_trace, proportion = 0.95)
|
Add the following code to your website.
For more information on customizing the embed code, read Embedding Snippets.