[R-sig-ME] MCMCglmm predict failing with more than one fixed effect

Tom Houslay hou@|@y @end|ng |rom gm@||@com
Fri Sep 27 17:11:25 CEST 2019

Hi all,

Just closing the open question I had in case anyone else has had the same
issue using predict.MCMCglmm. Turns out that it didn't seem to like
predicting with a single value in any of the fixed effect terms - so even
if I just want to predict for a single value of something, I just have to
add in another value and then ignore those rows afterwards. I've appended a
little bit to the reproducible code from my last email to show how that's
working (see below).




sleepstudy <- lme4::sleepstudy

# Basic model of sleepstudy with 1 fixed effect and simple random intercept
mcmc_ss_1 <- MCMCglmm(Reaction ~ Days,
                      random =~ Subject,
                      data = sleepstudy)


# Predicts fine on original data

# Create data on which to predict
df_ss_pred <- data.frame(Days = 0:9,
                         Subject = 350,
                         Reaction = 0)

# This prediction works
        newdata = df_ss_pred)

# Create new data frame with fake additional column
sleepstudy_2 <- cbind(sleepstudy,
                      extra = rnorm(nrow(sleepstudy)))

# Run model with additional fixed effect
mcmc_ss_2 <- MCMCglmm(Reaction ~ Days + extra,
                      random =~ Subject,
                      data = sleepstudy_2)


# Predict works fine from original data

# Create data frame on which to predict,
#  as before but now with new column added
df_ss_pred_2 <- as.data.frame(expand.grid(Reaction = 0,
                                          Days = 0:9,
                                          Subject = 350,
                                          extra = mean(sleepstudy_2$extra)))

# Prediction fails
        newdata = df_ss_pred_2)

# Check model formula

# Check variable names in data frame

# Check variable names in new data frame

# Try adding another value for 'extra'
df_ss_pred_3 <- as.data.frame(expand.grid(Reaction = 0,
                                          Days = 0:9,
                                          Subject = 350,
                                          extra =

mean(sleepstudy_2$extra) + sd(sleepstudy_2$extra))))

# Prediction now works!
        newdata = df_ss_pred_3)

	[[alternative HTML version deleted]]

More information about the R-sig-mixed-models mailing list