Chapter 15: solution/R/

Welcome to Chapter 15! In the previous chapter, 9-Real-World, we finished our journey through the core Machine Learning curriculum. We learned how to build, deploy, and use models in the real world using Python.

But Python isn't the only language spoken in the land of Data Science.

Just like in the human world, where you might say "Hello" in English or "Hola" in Spanish, in Data Science, you can write code in Python or R.

This brings us to the folder solution/R/.

Motivation: The Second Language

Imagine you are a chef.

• The Goal: Make a delicious soufflé.
• The Recipe: You have a recipe written in English (Python).
• The Problem: Your new sous-chef only speaks French (R).
• The Solution: You don't change the cooking process (the math is the same), but you translate the instructions.

The solution/R/ directory contains the exact same lessons you just learned, but translated into the R Programming Language.

Why does this exist? While Python is great for general programming (building apps, websites, and scripts), R was invented specifically by statisticians for statistics. If you go into academic research or specialized data analysis, you will see R everywhere.

Key Concepts: R vs. Python

If you open this folder, you will see files that look slightly different from what we used before.

1. R Markdown (.Rmd)

In notebook.ipynb, we used Jupyter Notebooks. The R equivalent is R Markdown.

• Python: Uses cells in a web interface.
• R: Uses a text file that "knits" code and text together into a beautiful report (PDF or HTML).

2. Assignment (<-)

This is the first thing that confuses beginners.

• Python: x = 5
• R: x <- 5 (It looks like an arrow, pointing the number 5 into the variable x).

3. The Tidyverse

In Python, we used Pandas to manipulate data. In R, we use a collection of tools called the Tidyverse. It is designed to be very easy to read, almost like English sentences.

How to Use This Abstraction

To use the files in solution/R/, you typically need a specific tool called RStudio. It is the "Command Center" for R, similar to how VS Code is for Python.

Step 1: Loading Data (Comparison)

Let's see how the code looks different for the exact same task: Loading our Pumpkin Data.

Python Version (What you know):

import pandas as pd
df = pd.read_csv("pumpkins.csv")
print(df.head())

R Version (What is in this folder):

library(tidyverse)

# The arrow puts data into 'df'
df <- read_csv("pumpkins.csv")

# Show the first few rows
head(df)

Explanation: They look almost identical! read_csv does the same thing in both languages. The main difference is the arrow <-.

Step 2: Plotting Data

R is famous for its beautiful graphs using a tool called ggplot2.

# Start a plot using the pumpkin data
ggplot(data = df, aes(x = Size, y = Price)) + 
  
  # Draw dots (points)
  geom_point()

Output: A scatter plot appears showing the relationship between Size and Price.

Explanation:

• aes: Aesthetics. We map the X axis to Size and Y axis to Price.
• geom_point: Geometric Object. We tell R to draw points (dots).

The Internal Structure: Under the Hood

How does the project organize these two languages? They run on parallel tracks.

When you look at the solution/R/ folder, you will see subfolders that match the main chapters perfectly.

Breakdown of the Map

• If you liked 2-Regression, look in solution/R/regression/.
• If you liked 4-Classification, look in solution/R/classification/.

Deep Dive: Solving Regression in R

Let's revisit the Pumpkin Price prediction problem from Chapter 7, but solve it the "Statistician's Way" using R.

In Python, we used Scikit-learn. In R, regression is built into the core language.

# 1. Train the model (Linear Model = lm)
# Formula: "Price depends on Size" (Price ~ Size)
model <- lm(Price ~ Size, data = df)

# 2. Look at the results
summary(model)

Output: R produces a very detailed statistical report, showing "P-values," "Residuals," and "F-statistics."

Explanation:

• lm(): Linear Model.
• ~ (Tilde): This symbol means "Depends on." We are saying Price depends on Size.
• summary(): Unlike Python's .predict(), R focuses heavily on explaining why the model works using statistics.

Making a Prediction

Just like in Python, we can ask our R model to guess the future.

# Create a new pumpkin of size 450
new_pumpkin <- data.frame(Size = 450)

# Predict the price
predict(model, new_pumpkin)

Output:

       1 
8.504 

Explanation: The result is $8.50, just like our Python model! Math is universal, even if the programming language changes.

Why this matters for Beginners

You might ask, "I just spent 14 chapters learning Python. Why should I care about R?"

1. Bilingualism: Knowing two coding languages makes you much more hireable.
2. Visualization: Many experts believe R's ggplot2 creates better, publication-quality graphs than Python's tools.
3. Specific Jobs: If you want to work in Biology, Psychology, or rigorous Statistics, R is often the preferred language.

Conclusion

In this chapter, we explored solution/R/. We learned that:

• It's a mirror: This folder contains the same lessons as the main course, translated into R.
• R Markdown: We use .Rmd files instead of .ipynb.
• Syntax: We use arrows <- and tildes ~, but the logic remains the same.

We have now seen how to do Machine Learning in English (Python) and French (R). But what if you want to read this tutorial in actual human languages, like Spanish, Chinese, or Hindi?

Global knowledge should be accessible to everyone, regardless of what language they speak.

Next Chapter: translations

Generated by Code IQ