survdistr 0.0.3
- Improved documentation
- Refactored
interp_cif()
- Added trapezoidal integration option for continuous hazard/density
to survival conversion in
convert_to_surv()
- Renamed argument:
trim_duplicates to
trim_dups
survdistr 0.0.2
- Added
convert_to_surv() for efficient transformation of
hazard/density functions to survival at anchor time points (uses C++
internally).
- Introduced
trim_duplicates() to remove repeated S(t)
values across the time axis, with tolerance.
- Added
interp_cif() for constant interpolation of CIF
(uses C++ internally).
- Added
extract_times() to consistently obtain and
validate time points from vectors or matrices.
- Unified assertion functions into one:
assert_prob().
- Unified
mat_interp() and vec_interp() into
a single interp() function.
- Enhanced
interp():
- Added aliases for the three interpolation options (e.g.,
const_haz is equivalent to exp_surv).
- Implemented S(t) interpolation in C++ for all 3 options
(
output = "surv"); H(t) and F(t) follow similarly.
- Implemented f(t) interpolation in C++ for all 3 options
(
output = "density")
- Implemented h(t) interpolation in C++ for all 3 options
(
output = "hazard")
- Requires unique and ordered
eval_times, resulting in
significant C++ speed-up.
- Allows passing
times separately to preserve
precision.
- Improved
survDistr:
- Removed
data_type argument; input now only S(t).
- Added aliases for the three interpolation options.
- Allows passing
times during construction separately for
precision.
- Introduced
filter() method for in-place object
filtering.
- Added
rows argument to filter input matrix before
interpolation without altering the object.
method and times are now active fields,
not public members.
- Added S3 conversion via
as_survDistr(x).
- Added unit tests for all features.
survdistr 0.0.1
- Base
survDistr class.
assert_prob_matrix() for asserting survival, CDF and
CIF matrices.
mat_interp() and vec_interp() for constant
and linear interpolation of survival, CDF and CIF matrices/vectors
(using internal Rcpp methods).