Overrides, queries, or resets the runtime policy consulted by
edi_warm_start_dispatch_policy() for whether an inference class
reuses a previous fit's parameters/curvature to seed the next fit during
resampling; see get_warm_start_dispatch_policy for the
built-in default table and its jackknife/non_param_boot/
bayesian_boot/param_boot/rand operation schema
(each with an inference_class_overrides layer and an
n_conditioned_overrides layer).
Value
Invisible NULL when policy is supplied (a mutation), or
invisibly the current policy configuration list when called for its
side-effect-free query/reset value.
Details
Call with no arguments (policy = NULL, reset = FALSE)
to retrieve the current configuration without changing it. Pass a named
list to policy to merge new/overriding entries into the current
configuration via modifyList (per-operation
sub-lists, e.g. list(rand = list(inference_class_overrides = ...))
or list(rand = list(n_conditioned_overrides = ...)), are merged
rather than replaced wholesale — note n_conditioned_overrides is a
plain list of rules, so overriding it replaces the whole list for that
operation, not a per-rule merge). Pass reset = TRUE to discard any
accumulated overrides and restore the package's built-in default policy
exactly as returned by get_warm_start_dispatch_policy.
This function controls the full dispatch policy, including the
sample-size-conditioned n_conditioned_overrides layer.
See also
get_warm_start_dispatch_policy for the policy schema
and built-in defaults; set_cold_start_dispatch_policy for
the analogous, simpler single-layer setter governing the initial
cold-start heuristic; tune_EDI_for_this_machine, which
calls this setter with machine-measured overrides rather than
hand-picked ones.