Thinking in Vectors, Not Loops

Thinking in Vectors, Not Loops

R doesn’t iterate — it transforms.
That’s the first truth to swallow if you want to stop fighting R and start flowing with it.

---

The Loop Mindset

Most programmers come to R from Python, C, or Java — languages that live on loops.
You take a list, run a loop, mutate an element, and build results piece by piece.
That mindset works there — but it breaks R’s rhythm.

# Python style
numbers = [2, 3, 5, 7, 11]
squares = []
for n in numbers:
    squares.append(n**2)
print(squares)

In Python, you describe steps.
In R, you describe relations.

---

The Vector Mindset

A vector is not a loop — it’s a container of intent.
When you square a vector, you’re not saying “for each element, square it.”
You’re saying:

Apply this operation to the entire idea of the data.

Dukungan Sistem Akademik

Kesulitan dengan riset atau tugas akademik Anda?

Tim ahli Notivra siap mendampingi Anda memberikan solusi bimbingan dan dukungan akademik yang komprehensif.

Konsultasi Sekarang

numbers <- c(2, 3, 5, 7, 11)
squares <- numbers^2
squares

That’s it.
No iteration, no indexing, no mutation.
Just transformation — clean, readable, fast.

---

Why This Matters

Vectorization isn’t just about performance.
It’s a mental shift from how to what.

When you think in vectors:

• You focus on relationships, not procedures.
• You express logic declaratively.
• Your code mirrors your reasoning — not your control flow.

That’s why R feels more like math than engineering.
You’re describing a system of transformations, not writing machinery.

---

Building Intuition

Let’s stretch it a bit.

temps <- c(29.4, 30.1, 31.6, 33.2, 32.0)
above_31 <- temps > 31
temps[above_31]

This tiny example carries big weight.
You’ve just filtered a dataset using logic as a vector — not a condition inside a loop.
This is R’s quiet genius: logical, numeric, and character vectors are all the same kind of object, and you can use one to filter another.

You can’t think that way in most languages — but in R, it’s the natural order.

---

The Tidyverse Connection

The entire tidyverse rests on this principle.
Functions like mutate(), filter(), summarize() — they’re all vectorized transformations.
Each one describes what should happen to columns of data, not how to iterate through them.

library(dplyr)

data.frame(x = 1:5, y = 6:10) |>
  mutate(sum_xy = x + y)

Every column is a vector.
Every operation is a transformation.
That’s why tidyverse code reads like poetry — concise, declarative, expressive.

---

Exercises

1. Create a vector of 20 random integers using sample(1:100, 20) and compute their cube roots without using loops.
2. Given scores <- c(85, 92, 77, 63, 95, 88), write one line to return all scores ≥ 90.
3. In your own data analysis scripts, find one for loop you can rewrite as a vectorized expression.

---

When you think in vectors, you stop commanding data — you start conversing with it.