rank_regression()
: For all distributions, the confidence intervals of the parameters are now computed on the basis of a heteroscedasticity-consistent (HC) covariance matrix. If the confidence intervals for the Weibull parameters are to be calculated according to Mock, this must be specified via the new argumentoptions
.mixmod_regression()
: Since this function usesrank_regression()
, the changes regarding the confidence intervals also apply here.plot_prob.wt_model()
: Removed dysfunctional argumentdistribution
. The distribution is inferred using the modelx
.
delta_method()
: Renamed argumentp
withx
.confint_betabinom()
andconfint_fisher()
: Removed constant featuresdistribution
,bounds
anddirection
from the tibble output and added them as attributes instead.
mcs_mileage()
: Changed name of output columnmileage
tox
(in accordance withreliability_data()
).mcs_delay()
: Changed name of output columntime
tox
(in accordance withreliability_data()
).dist_mileage.default()
(formerdist_mileage()
): Renamed argumentmileage
withx
.mcs_mileage.default()
(formermcs_mileage()
): Renamed argumentmileage
withx
.
- Implementation of one- and two-parametric exponential distribution (
'exponential'
and'exponential2'
).
estimate_cdf()
: Added optionjohnson_method
to specify the formula which is used for determining cumulative failure probabilities.
rank_regression()
: New argumentsdirection
(specifies direction of dependency in the model),control
(enables access to argumentcontrol
inoptim()
) andoptions
(method used to calculate the confidence intervals for the parameters, default is "HC").r_squared_profiling()
: New argumentdirection
.ml_estimation()
: New argumentsstart_dist_params
(optional vector with initial values of the parameters) andcontrol
(enables access to argumentcontrol
inoptim()
).loglik_profiling()
: New argumentwts
.loglik_profiling()
is now an S3 generic.loglik_profiling()
becomesloglik_profiling.default()
. Addedloglik_profiling.wt_reliability_data()
.loglik_function()
is now an S3 generic.loglik_function()
becomesloglik_function.default()
. Addedloglik_function.wt_reliability_data()
.
confint_betabinom()
: Methods"kaplan"
and"nelson"
ofestimate_cdf()
can be used for beta-binomial confidence bounds.
- Added
mcs_mileage_data()
: Create consistent MCS data formcs_mileage()
. - Added
mcs_delay_data()
: Create consistent MCS data formcs_delay()
. dist_mileage()
is now an S3 generic.dist_mileage()
becomesdist_mileage.default()
. Addeddist_mileage.wt_mcs_mileage_data()
.dist_delay()
is now an S3 generic.dist_delay()
becomesdist_delay.default()
. Addeddist_delay.wt_mcs_delay_data()
.dist_delay()
now supports the estimation of multiple delay distributions at once.mcs_mileage()
is now an S3 generic.mcs_mileage()
becomesmcs_mileage.default()
. Addedmcs_mileage.wt_mcs_mileage_data()
.mcs_delay()
is now an S3 generic.mcs_delay()
becomesmcs_delay.default()
. Addedmcs_delay.wt_mcs_delay_data()
.- Added
print.wt_mcs_delay_data()
andprint.wt_mcs_mileage_data()
. - Added
print.wt_mileage_estimation()
. - Added
print.wt_delay_estimation()
for one delay andprint.wt_delay_estimation_list()
for multiple delays.
- Fixed bug in
reliability_data()
: Using!!
syntax with argumentsx
andstatus
resulted in an error. estimate_cdf()
preserves additional columns, that were returned fromreliability_data(..., .keep_all = TRUE)
.- Improved
print.wt_reliability_data()
.
- Fixed bug in
plot_conf()
: Wrong confidence bounds were displayed fordirection = "x"
(#181). - Fixed bug in
plot_conf()
:plot_method = "ggplot2"
and exactly one method inestimate_cdf()
resulted in an error (#182).
- The object returned by
mcs_mileage()
now has classwt_mcs_mileage
. - The object returned by
mcs_delay()
now has classwt_mcs_delay
. - The object returned by
dist_mileage()
now has classwt_mileage_estimation
. - The object returned by
dist_delay()
now has classwt_delay_estimation
orwt_delay_estimation_list
.
plot_prob()
: Better work out the distinction betweenplot_prob.wt_cdf_estimation()
andplot_prob.wt_model()
. The former is applied to a CDF estimation whereas the latter is applied to a mixture model.
- Package now depends on R(>= 3.5.0)
mr_method()
: Deprecated, useestimate_cdf()
instead. Renamed output columncharacteristic
withx
. Set default value forid
toNULL
.johnson_method()
: Deprecated, useestimate_cdf()
instead. Renamed output columncharacteristic
withx
. Set default value forid
toNULL
.kaplan_method()
: Deprecated, useestimate_cdf()
instead. Renamed output columncharacteristic
withx
. Set default value forid
toNULL
.nelson_method()
: Deprecated, useestimate_cdf()
instead. Renamed output columncharacteristic
withx
. Set default value forid
toNULL
.plot_prob.default()
(formerplot_prob()
): Renamedevent
withstatus
.plot_prob_mix()
: Deprecated, useplot_prob()
instead. Removed default valueNULL
for argumentmix_output
. Renamedevent
withstatus
.
ml_estimation.default()
(formerml_estimation()
): Renamedevent
withstatus
. Removeddetails
. Changed names and contents of list elements in output. See?ml_estimation
.loglik_function
: Renamedevent
withstatus
. Renamedpars
withdist_params
.rank_regression.default()
(formerrank_regression()
): Renamedevent
withstatus
. Removeddetails
. Changed names and contents of list elements in output. See?rank_regression
.mixmod_em.default()
(formermixmod_em()
): Renamedevent
withstatus
. Removedpost
.mixmod_regression.default()
(formermixmod_regression()
): Renamedevent
withstatus
. Added argumentsk
andcontrol
, which provide finer control over the segmentation process. Expect default setting to provide other results than in prior versions.predict_prob()
: Renamedloc_sc_params
withdist_params
.predict_quantile()
: Renamedloc_sc_params
withdist_params
.plot_mod.default()
(formerplot_mod()
): Renamedevent
withstatus
. Renamedloc_sc_params
withdist_params
. Removedy
.plot_mod_mix()
: Deprecated, useplot_mod()
instead. Renamedevent
withstatus
.plot_pop()
: Added argumenttol
to restrict the range of failure probabilities. Removed argumentcolor
. Renamed argumentparams
todist_params_tbl
, which only supports location and scale parameters (also fordistribution = "weibull"
). Changed behavior ofdist_params_tbl
: Atibble
is now recommended instead of a vector.
confint_betabinom.default()
(formerconfint_betabinom()
): Renamedevent
withstatus
. Renamedloc_sc_params
withdist_params
. Added argumentb_lives
which allows the user to specify probabilitiesp
forB_p-lives
to be considered.confint_fisher.default()
(formerconfint_fisher()
): Renamedevent
withstatus
. Renamedloc_sc_params
withdist_params
. Renamedloc_sc_varcov
withdist_varcov
. Added argumentb_lives
which allows the user to specify probabilitiesp
forB_p-lives
to be considered.delta_method()
: Renamedloc_sc_params
withdist_params
. Renamedloc_sc_varcov
withdist_varcov
.plot_conf.default()
(formerplot_conf()
): Switched position of argumentsdirection
anddistribution
.
dist_delay_register()
: Deprecated, usedist_delay()
instead.dist_delay_report()
: Deprecated, usedist_delay()
instead.mcs_delay_register()
: Deprecated, usemcs_delay()
instead. Renamedx
withtime
. Renamedevent
withstatus
. Removedseed
. Removedint_seed
from output list.mcs_delay_report()
: Deprecated, usemcs_delay()
instead. Renamedx
withtime
. Renamedevent
withstatus
. Removedseed
. Removedint_seed
from output list.mcs_delays()
: Deprecated, usemcs_delay()
instead. Renamedx
withtime
. Renamedevent
withstatus
. Removedseed
. Removedint_seed
from output list.dist_mileage()
: Removedevent
. Renamedx
withtime
. Switched position of argumentstime
andmileage
.mcs_mileage()
: Removedevent
. Renamedx
withtime
. Switched position of argumentstime
andmileage
.
- Added support for ggplot2 in all plot functions. Plot method can be selected in
plot_prob()
orplot_pop()
via argumentplot_method
. - Added
reliability_data()
: Create consistent reliability data. - Added
estimate_cdf()
: Unite functionality ofmr_method()
,johnson_method()
,kaplan_method()
andnelson_method()
. Added optionties.method
formethod = "mr"
, which specifies how ties should be treated. - Support of multiple methods in
estimate_cdf()
and all functions that depend on thecdf_estimation
(rank_regression()
,plot_prob()
,plot_mod()
,plot_conf()
,mixmod_regression()
). plot_prob()
andplot_mod()
are able to handle mixture models.mixmod_regression()
is now more flexible. Argumentk
can be used to control number of subgroups or to determine them in an automatic fashion. Argumentcontrol
provides additional control over the segmentation procedure.- Added
print.wt_rank_regression()
,print.wt_ml_estimation()
,print.wt_model_estimation()
,print.wt_model_estimation_list()
,print.wt_mixmod_regression()
andprint.wt_mixmod_regression_list()
. - Added
vcov.wt_model_estimation()
. - Added
dist_delay()
: Generalizes the distribution-specific modeling of delays. - Added
mcs_delay()
: Generalizes the adjustment of operating times by delays and supports multiple delays at once. - Added lifecycle badges
rank_regression()
is now an S3 generic.rank_regression()
becomesrank_regression.default()
. Addedrank_regression.wt_cdf_estimation()
.plot_prob()
is now an S3 generic.plot_prob()
becomesplot_prob.default()
. Addedplot_prob.wt_cdf_estimation()
andplot_prob.wt_model()
.plot_mod()
is now an S3 generic.plot_mod()
becomesplot_mod.default()
. Addedplot_mod.wt_model()
.plot_conf()
is now an S3 generic.plot_conf()
becomesplot_conf.default()
. Addedplot_conf.wt_confint()
.plot_pop()
: Added support for multiple population lines and comparison of two- and three-parametric distributions.
- Revised README.
- Revised vignettes.
- Capitalized parameter documentation.
dist_delay_register()
anddist_delay_report()
: Usedist_delay()
instead.mcs_delay_register()
,mcs_delay_report()
andmcs_delays()
: Usemcs_delay()
instead.mr_method()
,johnson_method()
,kaplan_method()
andnelson_method()
: Useestimate_cdf()
instead.plot_prob_mix()
: Useplot_prob()
instead.plot_mod_mix()
: Useplot_mod()
instead.
calculate_ranks
.mixture_em_cpp
.plot_layout
.
- Fixed bug inside
plot_mod_mix()
for the case of no mixture distribution. - Fixed bug inside
confint_betabinom()
: many cases near one ->unique()
. - Fixed bug inside
mr_method()
: assigning a rank for the same lifetime. - Fixed bug inside
mixmod_regression
: call tosegmented::segmented.lm()
was incorrect. - Added trace type
"scatter"
and scatter mode"markers"
to plotly plots. delta_method()
,r_squared_profiling()
andloglik_profiling()
were vectorized.
- Fixed installation error when using clang compiler
- Since RcppArmadillo is used, the R version should be at least 3.3.0 (listed under Depends in DESCRIPTION file)
- Vignettes for non-parametric probability estimation, parameter estimation using Median-Rank Regression and Maximum-Likelihood and mixture model estimation are provided.
- Argument y in functions
plot_prob_mix()
andplot_mod_mix()
is deprecated and not used anymore. - Argument reg_output in functions
plot_prob_mix()
andplot_mod_mix()
is deprecated; use mix_output instead. - Function
plot_mod_mix()
was revised and updated in the way that the obtained results of the functionmixmod_em()
can be visualized. - Function
plot_prob_mix()
was revised and updated in the way that the obtained results of the functionmixmod_em()
can be visualized. - Implementation of EM-Algorithm using Newton-Raphson. The algorithm is written in c++ (
mixture_em_cpp()
) and is called inmixmod_em()
. - New method for the computation of Fisher's Confidence Bounds regarding probabilities is used. These method is called "z-Procedure" and is more appropriate to manage the bend-back behavior. Therefore an adjustment of functions
delta_method()
andconfint_fisher()
was made. - Implementation of log-location-scale models with threshold parameter like three-parametric Weibull ("weibull3"), three-parametric lognormal ("lognormal3") and three-parametric loglogistic ("loglogistic3").
- Implementation of location-scale models like smallest extreme value ("sev"), normal ("normal") and logistic ("logistic").
- Implementation of Log-Likelihood Profiling for three-parametric models in function
loglik_profiling()
. In general this function is used insideml_estimation()
for the purpose of estimating threshold parameter of three-parametric models. - Implementation of R-Squared Profiling for three-parametric models in function
r_squared_profiling()
. In general this function is used insiderank_regression()
for the purpose of estimating threshold parameter of three-parametric models. - Implementation of Log-Likelihood Function for all implemented models in function
loglik_function()
. In general this function is used insideml_estimation()
for the purpose of estimating the variance-covariance matrix of location-scale models "sev", "normal" and "logistic". The function is also used to estimate the variance-covariance matrix of log-location-scale models with a threshold parameter, i.e. "weibull3", "lognormal3" and "loglogistic3". - new argument in function
ml_estimation()
:wts
for case weights.