untransform_predictions: Untransform Vector of State Space Predictions

In luke-a-rogers/pbsedm: An R package to implement some of the methods of empirical dynamic modelling

Untransform Vector of State Space Predictions

Description

So this is the reverse of 'state_space_reconstruction_for_sve()', but we only need to do it for the response variable, so just a single vector. Was using the notation of in Appendix of first manuscript, but only one vector so changing to 'response_' to match the call from 'single_view_embedding_for_sve()'.

Usage

untransform_predictions(
  response_observed,
  response_s_predicted,
  max_lag,
  positive_response_only = TRUE
)

Arguments

response_observed	[numeric()][vector()] original observed values
response_s_predicted	[numeric()][vector()] state space predictions (from the state space that used lagged and scaled variables), though just the response column
max_lag	maximum lag used
positive_response_only	logical, if TRUE then set any negative predictions of the response variable to be the minimum observed value

Value

[numeric()][vector()] a prediction vector of original 'N_t' in absolute space, so 'response_predicted', but also with a value for 'T+1', which is what we are ultimately after.

luke-a-rogers/pbsedm documentation built on June 3, 2024, 5:20 a.m.