Impute missing samples using linear interpolation

interpolate_missing_samples(
  data,
  period = NULL,
  tol = 1e-06,
  interpolate_covariates = FALSE
)

Arguments

data	dataset to modify
period	period to check the times against (if `NULL`, first check to see if there is a known `period` set in the metadata, otherwise assumes 1)
tol	tolerance for the period
interpolate_covariates	`TRUE` or `FALSE` (whether to do covariates, too)

Value

the dataset, with interpolated samples

Details

First, check if the data are evenly sampled in time. If not, we exit early. Next, apply forecast::na.interp() to each variable that has non-finite values.