Impute missing values using non-commutative functions, i.e. where the order matters.
non-commutative imputation
Impute missing values using non-commutative functions, i.e. where the order matters.
na.cummax(.x, ...)na.cummin(.x, ...)
na.cumsum(.x, ...)
na.cumprod(.x, ...)
atomic-vector with 0 or more missing values
additional arguments
Non-commutative imputations functions assume that .x is in the proper order
since the values depend on order. Usually, this is relevant then .x is part
of a table.
These functions replaces NA values with the cummulative max of .x. Internally,
fun(.x, na.rm=TRUE, ... ) is used. If the function cannot be calculated
(e.g. .x isn't numeric) then x is returned unchanged with a warning.
Use of na.cumsum and na.cumprod are dangerous since they omit missing
values that may contribute to