Data as Objects, Analysis as Transformation

Data as Objects, Analysis as Transformation

In R, you don’t command the data — you transform it.

Every R session is a living stream. Data flows in, and with each pipe, it changes shape — cleaner, clearer, closer to insight.
That is the essence of R: data as objects, analysis as transformation.

---

🧩 Seeing Data as Objects

Everything in R is an object — data frames, vectors, lists, models.
You don’t just run commands; you manipulate structures.
This object orientation makes R a consistent and expressive tool.

# A tibble is more than a table — it’s an object with behavior
library(tibble)

df <- tibble(
  species = c("Elephant", "Tiger", "Turtle"),
  weight_kg = c(5400, 220, 90)
)

print(df)

Each column is a vector, each row a moment in your data’s story.

---

🔄 The Pipe: The Language of Flow

The pipe (|>, or %>% from magrittr) changed everything.
Instead of nesting functions, you think in transformations — one step flows into the next.

library(dplyr)

df |>
  mutate(weight_ton = weight_kg / 1000) |>
  filter(weight_ton > 0.1) |>
  arrange(desc(weight_ton))

Each line reads like a sentence: “Take the data, then mutate, then filter, then arrange.”
It’s not programming — it’s thought in motion.

---

🧠 Transformation as Understanding

Transformation is not just cleaning — it’s modeling your mental structure of reality.
When you reshape a dataset, you’re deciding what matters, what disappears, and what stays visible.
In this way, analysis becomes an act of philosophy: how you choose to see the world.

---

💡 Reflection

What does your pipeline say about how you think?

Try this:

library(dplyr)
starwars |>
  select(name, species, height, mass, homeworld) |>
  mutate(bmi = mass / (height/100)^2) |>
  arrange(desc(bmi)) |>
  head(5)

This is not just code — it’s a thought process, rendered in syntax.

---

📘 Try It Yourself

1. Take any dataset from datasets::
2. Apply at least three transformations using the pipe
3. Observe how your mental model of the data changes

---

🔗 Further Reading

• Wickham, H. (2014). Tidy Data. Journal of Statistical Software.
• Bache & Wickham (2014). Magrittr: A Forward-Pipe Operator for R.
• Hadley Wickham’s Advanced R, Chapter 6: Functions.

---

In R, transformation is not a step — it’s the soul of the analysis.