cut_ages().
This refactoring has resulted in minor tweaks to the given error
messages as well as a small bug fix relating to how
max_upper argument was being rounded (i.e. we were calling
round() rather than raising to the next integer(ish) number
with ceiling() as intended).Adds new functions reaggregate_counts() and
reaggregate_rates() which provide a simplified API for
reaggregation compared to reaggregate_interval_counts() and
reaggregate_interval_rates(). These use an elegant approach
to reaggregation suggested by @BlackEdder.
Deprecates (with warning of class “deprecatedWarning”) the following functions:
aggregate_age_counts()reaggregate_interval_counts()reaggregate_interval_ratessplit_interval_counts()The use of these could be a little confusing and it was hard to
achieve consistency with their APIs. reaggregate_counts()
and reaggregate_rates() can be used to replicate most of
the old functionality.
reaggregate_interval_rates() which
caused the first entry of the output to be incorrect when the
breaks were not specified from 0. Thanks to
@BlackEdder for
the report.aggregate_age_counts() will now only return a row
corresponding to NA ages if they were present in the input data.
Previously an NA-associated row would always be returned even if it’s
count was 0. Due to this change
reaggregate_interval_counts() will now never return an
NA-associated row.
split_interval_counts() now matches the
documentation and disallows missing (NA) bounds.
breaks_to_interval() and cut_ages()
both gain an argument, max_upper which allows users to
explicitly set the maximum upper bound.
New function reaggregate_interval_rates().
For the vignette we now use markdown as a lighter alternative to rmarkdown.
Initial release of ageutils which provides a collection
of efficient functions for working with individual ages and
corresponding intervals. These include functions for efficient
conversion from an age to an interval, aggregation of ages with
associated counts in to intervals and the splitting of interval counts
based on specified age distributions.
Functions are derived from those in the ympes with the intention being to remove these functions from that package in favour of this one going forward.