A blog for the things that don’t fit in a paper

meta
quarto
rstats

Starting a blog about R, statistics and data analysis: why, about what, and how this site is built.

Author

Ignacio Spiousas

Published

16 October 2026

There is a category of knowledge that accumulates as you work and never makes it into a paper: why that mixed model will not converge, how to actually read a periodogram, what scale() does to your coefficients, the least painful way to plot confidence intervals. This blog is for that.

What I will write about

  • R and the tidyverse — patterns I use every day, and the ones I abandoned.
  • Applied statistics — inference, mixed models, periodic data analysis, and the misunderstandings I see repeated in practice.
  • Visualisation — ggplot2, design decisions, and plots that communicate.
  • Workflow — reproducible projects, Quarto, version control.

The aim is for each post to be short, self-contained, and to carry code you can copy and run.

How this site is built

The site is written in Quarto and published with GitHub Pages. Every post is a .qmd file, so text and code live together and results are generated at render time:

library(tidyverse)

# An example of what tends to show up around here
tibble(t = seq(0, 4 * pi, length.out = 200)) |>
  mutate(signal = sin(t) + rnorm(n(), sd = 0.2)) |>
  ggplot(aes(t, signal)) +
  geom_line(linewidth = 0.4, colour = "grey60") +
  geom_smooth(method = "loess", formula = y ~ x, se = FALSE, colour = "#1d6a73") +
  labs(x = "Time", y = "Signal") +
  theme_minimal(base_size = 12)

One advantage of this setup: results are frozen (freeze: true), so the server that publishes the site does not need R installed. Code runs only when I change a post.

Coming up

I have a few notes queued on periodic data analysis, which is where I spend most of my time. If there is something you would like to read, write to me on GitHub or X.