add_lagged_covariates: Augment dataframe with lagged covariates

In simITS: Analysis via Simulation of Interrupted Time Series (ITS) Data

Description

Take outcome and a list of covariates and add new columns with lagged versions. Assumes rows of dataframe are in time ascending order. Lagged outcome canonically called 'lag.outcome'. Covariates 'lag.XXX'.

Usage

add_lagged_covariates(dat, outcomename, covariates = NULL)

Arguments

dat	The dataframe
outcomename	The outcome of interest (string)
covariates	The covariates to lag along with the outcome. This can be either of two things. First, it can be a list of string names. Covariates can also be a function with a "lags" attribute with the listed covariates (as returned by, e.g., make_fit_season_model) (which is a list of string names). NULL if no covariates other than outcome should be lagged.

Value

Augmented dataframe with lagged covariates as new columns. Will clobber old columns if the names (of form "lag.XXXX") conflict.

Examples

data( "newjersey" )
newjersey = add_lagged_covariates(newjersey, "n.warrant", c("sin.m","cos.m" ) )
head( newjersey[ c( "n.warrant", "sin.m", "lag.outcome", "lag.sin.m" ) ] )

simITS documentation built on July 2, 2020, 4:10 a.m.