Fill missing values

Fast fill missing values using constant value, last observation carried forward or next observation carried backward.

Usage

nafill(x, type=c("const", "locf", "nocb"), fill=NA, nan=NA)
setnafill(x, type=c("const", "locf", "nocb"), fill=NA, nan=NA, cols=seq_along(x))

Arguments

x

Vector, list, data.frame or data.table of numeric columns.

type

Character, one of "const", "locf" or "nocb". Defaults to "const".

fill

Numeric value to be used to replace missing observations. See examples.

nan

Either NaN or NA; if the former, NaN is treated as distinct from NA, otherwise, they are treated the same during replacement. See Examples.

cols

Numeric or character vector specifying columns to be updated.

Details

Only double and integer data types are currently supported.

Note that both nafill and setnafill provide some verbose output when getOption('datatable.verbose') is TRUE.

Value

A list except when the input is a vector in which case a vector is returned. For setnafill the input argument is returned, updated by reference.

See also

shift, data.table, fcoalesce

Examples

x = 1:10
x[c(1:2, 5:6, 9:10)] = NA
nafill(x, "locf")
#>  [1] NA NA  3  4  4  4  7  8  8  8

x = c(1, NA, NaN, 3, NaN, NA, 4)
nafill(x, "locf")
#> [1] 1 1 1 3 3 3 4
nafill(x, "locf", nan=NaN)
#> [1]   1   1 NaN   3 NaN NaN   4

# fill= applies to any leftover NA
nafill(c(NA, x), "locf")
#> [1] NA  1  1  1  3  3  3  4
nafill(c(NA, x), "locf", fill=0)
#> [1] 0 1 1 1 3 3 3 4

dt = data.table(v1=x, v2=shift(x)/2, v3=shift(x, -1L)/2)
nafill(dt, "nocb")
#> $v1
#> [1] 1 3 3 3 4 4 4
#> 
#> $v2
#> [1] 0.5 0.5 1.5 1.5 1.5  NA  NA
#> 
#> $v3
#> [1] 1.5 1.5 1.5 2.0 2.0 2.0  NA
#> 

setnafill(dt, "locf", cols=c("v2","v3"))
dt
#>       v1    v2    v3
#>    <num> <num> <num>
#> 1:     1    NA    NA
#> 2:    NA   0.5    NA
#> 3:   NaN   0.5   1.5
#> 4:     3   0.5   1.5
#> 5:   NaN   1.5   1.5
#> 6:    NA   1.5   2.0
#> 7:     4   1.5   2.0