Extrapolate Variables from Last Observation — extrapolate_from_last

Extrapolates variables from the last observation before a given time point using linear extrapolation with specified slopes.

Usage

extrapolate_from_last_observation(
  target_time,
  data,
  time_var,
  slopes = NULL,
  strict = TRUE
)

Arguments

target_time: Numeric scalar. The time point at which to extrapolate values.
data: A data.frame containing the observations, with one row per time point. Must be sorted by the time variable.
time_var: Character string. The name of the time variable in data.
slopes: A named numeric vector. Names should match column names in data for variables to extrapolate, and values are the slopes (rate of change per unit time). Variables not in slopes will be carried forward unchanged from the last observation.
strict: Logical. If TRUE (default), requires that target_time is greater than or equal to the last observed time. If FALSE, allows extrapolation to earlier times.

Value

A single-row data.frame with extrapolated values at target_time.

Details

The function finds the last observation in data where the time is less than or equal to target_time, then extrapolates variables using the formula:

$$x_{extrapolated} = x_{last} + (target\_time - time_{last}) \times slope$$

For variables not specified in slopes, the last observed value is used (slope = 0).

Examples

# \donttest{
# Create example data
df <- data.frame(
    time = c(0, 1, 2, 3),
    x1 = c(10, 12, 14, 16),
    x2 = c(5, 5.5, 6, 6.5),
    id = 1
)

# Extrapolate to time = 5 with known slopes
extrapolate_from_last_observation(
    target_time = 5,
    data = df,
    time_var = "time",
    slopes = c(x1 = 2, x2 = 0.5)
)
#>   time x1  x2 id
#> 4    5 20 7.5  1
# Expected: x1 = 16 + (5-3)*2 = 20, x2 = 6.5 + (5-3)*0.5 = 7.5

# Extrapolate with only some variables having slopes
extrapolate_from_last_observation(
    target_time = 4,
    data = df,
    time_var = "time",
    slopes = c(x1 = 2)
)
#>   time x1  x2 id
#> 4    4 18 6.5  1
# Expected: x1 = 16 + (4-3)*2 = 18, x2 = 6.5 (carried forward), id = 1
# }