NEWS

---

inferference 1.0.1

• Adds ability to fit some types of models with HLfit(family = binomial(), rand.family = gaussian()).

inferference 0.5.3

• fixes small issue with diagnose_weights() function
• fixes issue where weightd object was not showing up in output

inferference 0.5.2

• adds JSS citation

inferference 0.5.1

• updates vignettes to a stripped down version of the JSS paper
• effect utilities now only return a smaller subset of columns: alphas, treatments, estimate, std.error, and confidence limits

inferference 0.5.0

• Eliminates 'set_NA_to_0' to 'interference'.
• Modifies the 'integrate_allocation' argument to 'interference'. The default is now to include the allocation parameter in the weight term.
• Adds 'diagnose_weights' function which prints histograms of group-level weights.

inferference 0.4.62 (2015-06-08)

• To avoid confusion with the ‘method' argument of 'numDeriv::grad()', the 'causal_estimation_options' of 'interference()' has changed. 'variance_estimation = ’simple'‘ is now 'variance_estimation = ’naive''.