Classify new respondents with a fitted model

Assigns new respondents to the latent classes or ranks of a fitted model. The estimated parameters (and, for biclustering, the field memberships of the items) are held fixed and one E-step is run on the new responses, with the uniform class prior the estimation itself uses (or the rank prior given by trapezoidal in the ordinal and rated LRA). On the data the model was fitted to, the result reproduces the membership stored in the fit.

Supported fits: LCA() (binary, nominal, ordinal, rated), LRA() (binary, ordinal, rated) and Biclustering() / Ranklustering (binary, nominal, ordinal, rated). The IRM and network models are not supported.

Usage

# S3 method for class 'exametrika'
predict(object, newdata, na = NULL, ...)

Arguments

object

A fitted model.

newdata

New responses: a matrix or data frame with the items in columns (named as in the fitted data; extra columns are ignored, and the item names are required unless the column count already matches), or an exametrikaData object from dataFormat(). Missing responses (NA, or the codes given in na) are allowed and simply left out of the likelihood. Responses must use the coding of the fitted data.

na

Values to treat as missing, besides NA.

...

Not used.

Value

An object of class exametrikaPrediction, a list with

Membership

Matrix of posterior membership probabilities (respondents x classes or ranks).

Estimate

The class or rank with the largest membership.

log_lik

Log-likelihood of each respondent's responses under the model (the class or rank mixed out).

model

The class of the fitted object.

msg

"Class" or "Rank".

A response with a category the fitted data never used for that item cannot be scored by the model; it is treated as missing, with a warning.

Examples

# \donttest{
fit <- LCA(J15S500, ncls = 5)
pred <- predict(fit, J15S500$U[1:10, ])
pred$Estimate
#>  Student1  Student2  Student3  Student4  Student5  Student6  Student7  Student8 
#>         1         3         3         4         1         4         5         4 
#>  Student9 Student10 
#>         5         2 
# }