expose_split()
can convert any exposed_df
object with calendar period exposures (yearly, quarterly, monthly, or weekly) into a split_exposed_df
object. Split exposure data frames contain columns for exposures both on a calendar period and policy year basis.exp_stats()
and exp_shiny()
now require clarification as to which exposure basis should be used when passed a split_exposed_df
object.expose_df
objects now contains a default_status
attribute.autotable()
functions now contain the arguments decimals_amt
and suffix_amt
. The former allows one to specify the number of decimals appearing after amount columns. The latter is used to automatically scale large numbers into by thousands, millions, billions, or trillions.exp_stats()
is passed a weighting variable.summary()
method for exposed_df
objects that calls exp_stats()
.expose()
functions was changed from the first observed status to the most common status.as_exp_df()
and as_trx_df()
were added to convert pre-aggregated experience studies to the exp_df
and trx_df
formats, respectively.agg_sim_dat
- a new simulated data set of pre-aggregated experience was added for testing as_exp_df()
and as_trx_df()
.is_exp_df()
and as_trx_df()
were added to test for the exp_df
and trx_df
classes.conf_int
argument was added to exp_stats()
that creates confidence intervals around observed termination rates, credibility-weighted termination rates, and any actual-to-expected ratios.conf_int
was added to trx_stats()
to create confidence intervals around utilization rates and any “percentage of” output columns. A conf_level
argument was also added to this function.autoplot.exp_df()
and autoplot.trx_df()
now have a conf_int_bars
argument that plots confidence intervals (if available) as error bars for the selected y-variableautoplot.exp_df()
and autoplot.trx_df()
can now create scatter plots if “points” is passed to the geoms
argument.autoplot()
methods was updated to use an area geometry instead of bars for discrete x-axis variables. In addition, when a log-10 y-scale is used, areas will always be positive quantities. Previously, it was observed that areas were drawn as negative values for y-values on the main scale less than 1.autotable.exp_df()
and autotable.trx_df()
were updated to format intervals.exp_shiny()
updates
Breaking change - The confidence level argument cred_p
was renamed to conf_level
. This change was made because the confidence level is no longer strictly used for credibility calculations. This change impacts the functions exp_stats()
and exp_shiny()
.
autoplot.exp_df()
and autoplot.trx_df()
now include new options for adding a second y-axis and plotting results on a log-10 scale. The second y-axis defaults to plotting exposures using an area geometry.autoplot()
methods. These include plot_termination_rates()
and plot_actual_to_expected()
for termination studies and plot_utilization_rates()
for transaction studiesexp_shiny()
function received a handful of updates to accommodate new plotting functions and options. A small performance improvement was added in filtering logic as well. New options include a title input, credibility options taken from exp_stats()
,add_predictions()
and step_expose()
.autoplot()
and autotable()
methods?actxps
)add_predictions()
function that attaches one or more columns of model predictions to an exposed_df
object or any other data frame.add_transactions()
and autotable()
functions for compatibility with the dplyr 1.1.1 and gt 0.9.0.The actxps package now contains support for transaction studies.
add_transactions()
function adds transactions to exposed_df
objects.trx_stats()
function summarizes transaction results and returns a trx_df
object.trx_df
) S3 methods were added for for autoplot()
and autotable()
.exp_shiny()
function was updated to support transaction studies.withdrawals
) and sample policy values (account_vals
). These are meant to be paired with census_dat
.vignette("transactions")
.Other changes
pol_interval()
(a generic version), pol_yr()
, pol_qtr()
, pol_mth()
, and pol_wk()
. See vignette("misc")
.as_exposed_df()
function to include stricter input requirements and helpful error messages.exposed_df
objects to ensure class persistence, especially on grouped data frames. These include: group_by()
and ungroup()
, filter()
, arrange()
, mutate()
, select()
, slice()
, rename()
, relocate()
, left_join()
, right_join()
, inner_join()
, full_join()
, semi_join()
, and anti_join()
.autotable.exp_df()
was updated to be consistent across like columns.pol_val
column in census_dat
was renamed to premium
.expose()
functions now include a new column for period end dates.
Fixed issues with expose()
dropping records:
Fixed 2 R CMD check problems.
First version submitted to CRAN.
Added exp_shiny()
function.
Added step_expose()
recipe step function.
First developmental version