Skip to contents

Interpolate missing values in grouped data

Source: R/interpolate_na.R
interpolate_na.Rd

Perform linear interpolation to fill missing (NA) values in specified columns (vars2), with grouping based on a unique row identifier (var1). Interpolation is performed row-wise. See Examples for the structure of input data.

Usage

interpolate_na(data, var1, vars2, maxgap = 1, rule = 1)

Arguments

data

A data frame containing the variables to be interpolated.

var1

A variable serving as the unique row identifier (group). Each value in var1 must be unique. If duplicate values exist, the function will return an error.

vars2

A vector specifying the numeric columns to interpolate.

maxgap

An integer defining the maximum number of consecutive NA values to interpolate. Defaults to 1. If gaps exceed this limit, the values remain NA.

rule

An integer (1 or 2) specifying the extrapolation rule. Defaults to 1, meaning no extrapolation beyond available data range. Rule 2 uses the nearest observed value for out-of-range points. A vector (e.g., rule = c(2,1)) can specify different extrapolation rules for the left and right sides.

Value

A data frame with missing values interpolated in vars2.

Examples

df <- data.frame(
  date = c("2022-03-01", "2022-08-13", "2023-02-09", "2024-09-22"),
  temp_1am = c(NA, 20, 5, 24),
  temp_6am = c(6, NA, 6, NA),
  temp_11am = c(13, 25, 8, NA),
  temp_4pm = c(18, 30, NA, 30)
)

interpolate_na(df,
               var1 = "date",
               vars2 = c("temp_1am", "temp_6am", "temp_11am", "temp_4pm"))
#> # A tibble: 4 × 5
#>   date       temp_1am temp_6am temp_11am temp_4pm
#>   <chr>         <dbl>    <dbl>     <dbl>    <dbl>
#> 1 2022-03-01       NA      6          13       18
#> 2 2022-08-13       20     22.5        25       30
#> 3 2023-02-09        5      6           8       NA
#> 4 2024-09-22       24     NA          NA       30

interpolate_na(df,
               var1 = "date",
               vars2 = c("temp_1am", "temp_6am", "temp_11am", "temp_4pm"),
               rule = 2)
#> # A tibble: 4 × 5
#>   date       temp_1am temp_6am temp_11am temp_4pm
#>   <chr>         <dbl>    <dbl>     <dbl>    <dbl>
#> 1 2022-03-01        6      6          13       18
#> 2 2022-08-13       20     22.5        25       30
#> 3 2023-02-09        5      6           8        8
#> 4 2024-09-22       24     NA          NA       30