How to make a function of MAE and RAE without using library(Metrics)?

Question

My goal is to create functions of Mean Absolute Error (MAE) and Relative Absolute Error (RAE) without using any kind of library, like library(Metrics).

I try to input a formula inside a function for both MAE and RAE

mae <- function(a, b)
{
    mean(abs(a, b))
}


rae <- function(a, b)
{
    abs(a,b )
}

However, both function provide different answer compare to the function below,

library(Metrics)
stats <- function(a,b)
    {
        mae <- mae(a,b)
        rae <- rmse(a,b)
    }

Answer 1

Your function is wrong. You can use the following function to calculate mae and rae without any package

x <- c(1.1, 1.9, 3.0, 4.4, 5.0, 5.6)
y <- c(0.9, 1.8, 2.5, 4.5, 5.0, 6.2)

mae1 <- function(x,y)
  {
  mean(abs(x-y))
  }

mae1(x, y)
#> [1] 0.25


rae1 <- function(x,y)
  {
  sum(abs(x-y))/sum(abs(x - mean(x)))
  }

rae1(x,y)
#> [1] 0.1666667

library(Metrics)
#> Warning: package 'Metrics' was built under R version 3.6.2

mae(x, y)
#> [1] 0.25
rae(x,y)
#> [1] 0.1666667

^{Created on 2019-12-27 by the reprex package (v0.3.0)}

It is giving same output as that of Metrics package.

Update

If your data contains NAs then the above functions, as well as the functions of package Matrics, will fail. Under such situation use the following code

x <- c(1.1, 1.9, 3.0, 4.4, 5.0, 5.6, NA)
y <- c(0.9, 1.8, 2.5, 4.5, 5.0, 6.2, 2)

mae1 <- function(x,y, na.rm=TRUE)
  {
  mean(abs(x-y), na.rm=na.rm)
  }

mae1(x, y, na.rm=TRUE)
#> [1] 0.25


rae1 <- function(x,y, na.rm=TRUE)
  {
  sum(abs(x-y), na.rm=na.rm)/sum(abs(x - mean(x, na.rm=na.rm)), na.rm=na.rm)
  }

rae1(x,y, na.rm=TRUE)
#> [1] 0.1666667

library(Metrics)
#> Warning: package 'Metrics' was built under R version 3.6.2

mae(x, y)
#> [1] NA
rae(x,y)
#> [1] NA

^{Created on 2019-12-27 by the reprex package (v0.3.0)}