tipr 1.0.2
CRAN release: 2024-02-06
-
adjust_coef_with_binary()now assumes the coefficient is from a linear model rather than loglinear. Useloglinear = TRUEto get the old behavior. (#12, @malcolmbarrett) - Fixed roxygen issue with package documentation
- Update messaging and errors
tipr 1.0.1
CRAN release: 2022-09-05
- Fixed bug, functions based on the
adjust_coef_with_binaryfunction had the old parameter names (exposed_pandunexposed_p). These were changed to match the other new updates from version 1.0.0 to now beexposed_confounder_prevandunexposed_confounder_prev. - Change “relative risk” to “risk ratio” in all documentation.
- Add new JOSS citation
tipr 1.0.0
CRAN release: 2022-08-06
Breaking changes. The names of several arguments were changed for increased clarity:
effect->effect_observedoutcome_association->confounder_outcome_effectsmd->exposure_confounder_effectexposed_p->exposed_confounder_prevunexposed_p->unexposed_confounder_prevexposure_r2->confounder_exposure_r2outcome_r2->confounder_outcome_r2Added two new example datasets:
exdata_continuousandexdata_rr
tipr 0.4.1
CRAN release: 2022-05-05
- Add additional functions that specify
*_with_continuous()(long form of, the function names, the default unmeasured confounder is Normally distributed) - Change
tip_lm()totip_coef().
tipr 0.4.0
CRAN release: 2022-04-16
- Changed the name of
lm_tip()totip_lm() - The API has been fundamentally updated so that the functions now take a numeric value as a first argument rather than a data frame.
- Added adjust_* functions to allow for specification of all unmeasured confounder qualities without tipping
- Split
tip_*functions into hazard ratio, odds ratio, and relative risk - Add R2 parameterization with
tip_coef_with_r2(),adjust_coef_with_r2(), andr_value()
tipr 0.3.0
CRAN release: 2021-09-10
- Added ability to perform sensitivity analyses on linear models via
lm_tip()
tipr 0.2.0
CRAN release: 2020-11-16
- Updated several function and parameter names. The main functions are now
tip()andtip_with_binary(). The parameter names are more self-explanatory. - The API has been fundamentally updated so that the functions now take a data frame as a first argument.
- There is now explicit (but not required) integration with the
broompackage.
