estimate AFE/PAF from adjusted logistic regression models. CIs estimated from non-parametric bootstrap.
provides estimates of the excess cases in the exposed (with 95% CIs)
other improvements to estimates and output generally
parallel option to help speed up bootstrapping
yodr 1.1.0 (17th March 2026)
New features
phewas can now utilise more cores by running with parallel=TRUE
{parallel} package added to dependencies
New function estimate_ld to estimate haplotype frequencies and LD statistics between two biallelic variants
New function paf to estimate attributable fraction in the exposed (AFE) and population attributable fraction (PAF) with 95% CIs for a binary exposure (coded 0/1)
Can be a simple binary outcome (prevalence / risk) or a time-to-event outcome (incidence) and uses a Cox proportional hazards model
yodr 1.0.0 (22nd Feb 2026)
Changes
Package renamed to something 55% shorter but 100% better. Follows the general *nix rule that if a command is >4 characters it’s not worth typing. Also, yoda.
Renamed tidy_ci to tidy and get_assoc to est in the same spirit (couldn’t work out how to make these Star Wars themed whilst also being max 4 characters)
Together, this means where before I kept having to type lukesRlib::tidy_ci(fit) now I can type yodr::tidy(fit) and be happer
Original package lcpilling/lukesRlib still there and unchanged so all my old scripts still work