The hardware and bandwidth for this mirror is donated by METANET, the Webhosting and Full Service-Cloud Provider.
If you wish to report a bug, or if you are interested in having us mirror your free-software or open-source project, please feel free to contact us at mirror[@]metanet.ch.
cor_matrix() now draws column i
from distribution i (previously random with
replacement, so columns could repeat or omit inputs); columns are now
labelled with distribution names. Results will differ.[-1, 1] range,
unit diagonal) are now shared across mcs(),
smm(), and sensitivity().sensitivity() warns when an index is negative, which
can happen under negative correlation.cost_pdf() no longer requires
sum(risk_probs) <= 1.cost_post_pdf() gained risk_probs, so
unobserved (NA) risks are drawn from their prior instead of
treated as absent; warns if omitted with NAs present.predict_sigmoidal() now errors on a
model_type/fit mismatch instead of returning
NA confidence bounds.mcs() fixes the Cholesky decomposition for
correlated tasks — target correlations and marginal means were
previously distorted. Correlated results will differ; independent tasks
are unaffected.pra_chat(), pra_app(), slash commands, and the
RAG functions; use pra_mcp_server() and the bundled Agent
Skills (inst/skills/) instead.ragnar, shiny,
shinychat, bslib, vitals,
mockr; added corrplot, igraph,
networkD3, devtools, remotes,
mcptools.prob_net(), prob_net_sim(),
prob_net_learn(), prob_net_update().links are now load-bearing — edges must match the
distributions’ dependencies, nodes must be topologically ordered, and
adjacency_matrix is now directed. Networks that were
already self-consistent are unaffected.risk_prob()/risk_post_prob() combine multiple
causes with a noisy-OR, keeping results in [0, 1].
Single-cause results are unchanged.sensitivity() indices now reflect each task’s actual
variance contribution instead of always returning 1.0 for independent
tasks.building_project example dataset.inst/paper/pra-jss/).These binaries (installable software) and packages are in development.
They may not be fully stable and should be used with caution. We make no claims about them.