predict.sento_model: Make predictions from a sento_model object

In sborms/sentometrics: An Integrated Framework for Textual Sentiment Time Series Aggregation and Prediction

Description

Prediction method for sento_model class, with usage along the lines of predict.glmnet, but simplified in terms of parameters.

Usage

## S3 method for class 'sento_model'
predict(object, newx, type = "response", offset = NULL, ...)

Arguments

object	a sento_model object created with sento_model.
newx	a data matrix used for the prediction(s), row-by-row; see predict.glmnet. The number of columns should be equal to sum(sento_model$nVar), being the number of original sentiment measures and other variables. The variables discarded in the regression process are dealt with within this function, based on sento_model$discarded.
type	type of prediction required, a value from c("link", "response", "class"), see documentation for predict.glmnet.
offset	not used.
...	not used.

Value

A prediction output depending on the type argument.

Author(s)

Samuel Borms

See Also

predict.glmnet, sento_model

sborms/sentometrics documentation built on Aug. 21, 2021, 6:40 a.m.