Creating a custom metric • HelpR

This article provides example code showing how the MeasureR framework can be implemented to calculate a custom metric.

library(dplyr)
library(ggplot2)
library(ConnectR) # Will use example data from this package

E.g. Influence

Say that after working through the MeasureR framework, we decided that we thought:

\[\text{Influence} = \frac{1}{2}\text{Followers} + \frac{1}{2}\text{Reach}\]

with_metric <- retweet_example %>%
  # Standardise the variables we want to combine
  mutate(across(.cols = c(followers, `Reach (SUM)`),
                .fns = ~as.vector(scale(.)),
                .names = "{col}_scale")) %>%
  # Calculate a row-wise mean (for equal weighting) to define the new metric
  rowwise() %>%
  mutate(influence = mean(followers_scale, `Reach (SUM)_scale`))
  
# Check if there are any missing values in our custom metric
sum(is.na(with_metric$influence))

## [1] 0