Tests only: fixes the test failures on macOS reported in the CRAN checks of version 1.0.0. The package code is unchanged.
Consolidated release of the package, based on more than two years of use. New in this version: the influence of quality controls on the distribution of basic variables can be taken into account.
PROB_QUALCON. The conformity criterion of EN 206 for the compressive
strength ("fck_EN-206") and an attribute plan for dimensions and other
characteristics ("dim_k||n") are available. The acceptance probability of
a plan (OC curve) can be computed, plotted or entered manually.PROB_BASEVAR$updateBasevar() returns the basic variable after acceptance of
the lot (posterior distribution). The prior must be a Student-t ("st") or
log-Student-t ("lt") distribution with hyperparameters. The message compares
mean value and standard deviation of prior and posterior."MCIS") was revised. It estimates the
system failure probability more accurately and also handles parallel systems
whose design points are far apart.calculateSystemProbability(): "simpleBounds",
"MCIS", "MCC" (crude Monte Carlo) and "MCSUS" (subset simulation).
FORM and SORM provide single results, which enter the simple bounds. A system
analysis by FORM or SORM is not available.Name.MC_IS now counts the samples in the
failure domain.optim_type = "auto" (default) switches to a second solver if the
first one does not converge. The result field fallback shows whether this
happened.debug.level.PROB_BASEVAR$plot() plots the probability density of a basic variable. For a variable created by updateBasevar(), the prior distribution can be added with prior = TRUE and the fitted curve and the empirical resample with compare = TRUE.MC_IS_single to ensure correct accumulation of
weighted samples and consistent global statistics.sys_type from calculateSystemProbability() to MC_IS_system(),
ensuring that serial and parallel systems are evaluated correctly.mode_ids, shift, and references to I_matrix).MC_IS_single, including linear limit-state cases,
non-normal distributions, and rare-event scenarios.MC_CRUDE.SYS_PROB, SYS_PARAM, SYS_LSF, PROB_BASEVAR, …).Cov <- Inf) to reflect infinite relative scatter.func changed from "function" to "ANY" with explicit default NULL.
This prevents creation of dummy closures and allows proper missing-function detection.$check() method:
SYS_PARAM now correctly forwards arguments to parent class (callSuper(...))
ensuring that inherited fields like sys_input are preserved..onLoad().
The outer level is capped by environment variable TesiproV.max_workers
(default: 4 cores) while inner Monte-Carlo methods manage their own threads.$prepare():
location = Mean + digamma(1) * scale; scale derived via (Sd * sqrt(6)) / π).$getlDistr() for performance improvement..onLoad() function for package-specific default options that can later be used by TesiproV
functions to configure parallel execution in a controlled and CRAN-compliant way.$printResults().Existing scripts using older class definitions should continue to work after minor adjustments:
replace direct field assignments by constructor arguments where applicable,
and ensure that limit-state functions are defined before calling $check().
Better implementation of parallel computing in Unix and macOS environments.
Changed RNG handling in parallel approach.
© 2021–2026 T. Lux, K. Nille-Hauf, T. Feiri, M. Ricker -- Hochschule Biberach (until 2022), TU Dortmund University – Chair of Structural Concrete (since 2023)