MICE
@amices.bsky.social
Home of the #RStats imputation package {mice}. Amices is a place for people interested in solving missing data problems. See https://amices.org & https://github.com/amices. Posts by @oberman.bsky.social
What if you want to predict the missing values instead of doing inference with incomplete data?
Hanne Oberman@oberman.bsky.social · 3mo agoIf you're not interested in the uncertainty due to missingness, then just use a non-Bayesian (or even non-stochastic) @amices.bsky.social method, like "norm.predict" or "pmm" with only 1 donor
"mice: the only method here that carries imputation uncertainty into the model. Use it whenever the output is an estimate, a confidence interval, or a p-value, the typical medical or epidemiological analysis. Ten years on, it is still the right default for inference."
DataScience+@datascienceplus.bsky.social · 2w agoImputing Missing Data in R: mice, missRanger, and VIM Compared #rstats
Meet Mr. MICE himself! This summer, Stef van Buuren teaches 'Solving Missing Data Problems in R' at @utrechtuniversity.bsky.social. Learn cutting-edge techniques for addressing missing data problems in statistics and modern machine learning workflows. utrechtsummerschool.nl/courses/data...
Do you struggle with missing data in your statistical analyses? Come join our Summer School "Solving Missing Data Problems in R"! We will tell you when you can safely use complete case analysis, and when more dedicated strategies are required! utrechtsummerschool.nl/courses/data...
Cool stuff! Florian van Leeuwen and I implemented a prediction function in the #mice package that allows the incorporation of missing data uncertainty in a prediction interval. The `predict_mi()` function is available in the current development version: github.com/amices/mice #Rstats #statsky
![Image of R code. To reproduce:
library(ggplot2)
library(dplyr)
library(mice, warn.conflicts = FALSE)
imp <- mice(nhanes, m = 5, maxit = 5, seed = 1,
ignore = rep(c(FALSE, TRUE), c(20, 5)),
print = FALSE)
impdats <- complete(imp, "all")
train <- lapply(impdats, function(dat) subset(dat, !imp$ignore))
test <- lapply(impdats, function(dat) subset(dat, imp$ignore))
fits <- lapply(train, function(dat) lm(age ~ bmi + hyp + chl, data = dat))
preds <- predict_mi(object = fits, newdata = test, pool = TRUE, interval = "prediction")
preds
preds %>%
as.data.frame() %>%
mutate(case = 1:nrow(preds),
y = test[[1]]$age) %>%
ggplot(aes(x = fit, y = case, col = rowSums(is.na(nhanes[imp$ignore,]))>0)) +
geom_point() +
geom_errorbar(aes(xmin = lwr, xmax = upr)) +
theme_minimal() +
scale_color_manual(values = mice::mdc(1:2), labels = c("observed", "missing")) +
theme(legend.title = element_blank(),
legend.position = "bottom") +
labs(x = "prediction",
title = "Pooled prediction intervals")](https://cdn.bsky.app/img/feed_thumbnail/plain/did:plc:omgr6y6c62b4qar7fganjtdy/bafkreiesbac4qo25mh5rkegnw4ntcvbwedvpzqms4zn7xiothhzlgchxcm)
Is mean imputation effective? Yes—at filling in the missings; but also Yes—at biasing any subsequent analyses. stefvanbuuren.name/fimd
📊 Struggling with missing data in your analyses? Join our 4-day course 'Advanced Techniques for Handling Missing Data' at @utrechtuniversity.bsky.social! 📅 24–27 Mar 2025 📍 Utrecht, NL 💶 €730 | 1.5 ECTS Learn cutting-edge imputation with {mice} in R. Apply by March 10th! #RStats #DataScience
Advanced Techniques for Handling Missing Data in analysis and prediction workflows | Utrecht Summer School
This 4-day course provides cutting-edge techniques for addressing missing data problems, focusing on the intersection of statistical theory and modern machine learning workflows.
utrechtsummerschool.nl