[This article was first published on Seascapemodels, and kindly contributed to R-bloggers]. (You can report issue about the content on this page here)Want to share your content on R-bloggers? click here if you have a blog, or here if you don't. I’m a big fan of Github Copilots ghost text and next edit suggestions. As a data analyst they give you much more control over the pace and direction of analysis code than a full AI agent. They also let you practice the core skill an agent needs from you: writing a clear specification.I’ll walk through it in R, but the same ideas apply in any language. I’m assuming you’re on VScode with the Copilot extension installed. This should work with the free or paid plans (though I haven’t checked free myself).Start with ghost textGhost text is the greyed-out completion that appears as you type. Say you have a long-format dataset with timeseries for three species:library(dplyr)library(ggplot2)dat