lasso_2_scalar_forecast: Obtain scalar forecast from lasso model

In kassandra-ru/kassandr: Functions to Forecast Russian Macro Data in Kassandra Project

Description

Obtain scalar forecast from lasso model

Usage

lasso_2_scalar_forecast(
  model,
  h = 1,
  target = "value",
  model_sample,
  s = c("lambda.min", "lambda.1se")
)

Arguments

model	estimated lasso model
h	forecasting horizon
target	name of the target variable
model_sample	non-augmented data set for model estimation
s	criterion to select best regularization lambda in lasso

Details

Obtain scalar forecast from lasso model. The function automatically augments data with lags, fourier terms, trend etc.

Value

scalar forecast for given h

Examples

test_ts = stats::ts(rnorm(100), start = c(2000, 1), freq = 12)
test_tsibble = tsibble::as_tsibble(test_ts)
model = lasso_fun(test_tsibble, h = 1)
lasso_2_scalar_forecast(model, h = 1, model_sample = test_tsibble)

kassandra-ru/kassandr documentation built on Jan. 1, 2022, 7:39 a.m.