Create a new data frame for prediction

Source: R/predict.R

Build a data.frame for prediction, where one variable varies and all other variables are set to the reference value (median for continuous variables).

predDF(object, ...)

# S3 method for JointAI
predDF(object, vars, length = 100L, ...)

# S3 method for formula
predDF(object, data, vars, length = 100L, ...)

Arguments

object

object inheriting from class 'JointAI'
...

optional specification of the values used for some (or all) of the variables given in vars
vars

name of variable that should be varying
length

number of values used in the sequence when vars is continuous
data

a data.frame containing the original data (more details below)

See also

predict.JointAI, lme_imp, glm_imp, lm_imp

Examples

# fit a JointAI model
mod <- lm_imp(y ~ C1 + C2 + M2, data = wideDF, n.iter = 100)

# generate a data frame with varying "C2" and reference values for all other
# variables in the model
newDF <- predDF(mod, vars = ~ C2)

head(newDF)

#>           y       C1         C2 M2
#> 1 -3.183137 1.434298 -0.9226220  1
#> 2 -3.183137 1.434298 -0.9046136  1
#> 3 -3.183137 1.434298 -0.8866052  1
#> 4 -3.183137 1.434298 -0.8685968  1
#> 5 -3.183137 1.434298 -0.8505885  1
#> 6 -3.183137 1.434298 -0.8325801  1


newDF2 <- predDF(mod, vars = ~ C2 + M2,
                 C2 = seq(-0.5, 0.5, 0.25),
                 M2 = levels(wideDF$M2)[2:3])
newDF2

#>            y       C1    C2 M2
#> 1  -3.183137 1.434298 -0.50  2
#> 2  -3.183137 1.434298 -0.25  2
#> 3  -3.183137 1.434298  0.00  2
#> 4  -3.183137 1.434298  0.25  2
#> 5  -3.183137 1.434298  0.50  2
#> 6  -3.183137 1.434298 -0.50  3
#> 7  -3.183137 1.434298 -0.25  3
#> 8  -3.183137 1.434298  0.00  3
#> 9  -3.183137 1.434298  0.25  3
#> 10 -3.183137 1.434298  0.50  3