3. Building with AI: Build a Real Analysis in 30 Minutes

Put it all together: load data, clean it, summarize it, plot it.

Duration: 30 min

Tools: Cursor, R, RStudio or R Markdown

What You’ll Learn

How to use Cursor Chat to write real analysis code
Build a complete workflow from data → summary → visualization
Quick debugging tips when things go wrong

Cursor Shortcuts

Shortcut	What It Does
`Cmd+L`	Open Chat (ask questions, get code)
`Cmd+I`	Open Composer (generate full sections)
`Cmd+K`	Inline edit (fix selected code)

The Project: Quick Penguin Report

Create a new .R file and use Cursor Chat to build these 4 steps:

Step 1: Load & Inspect (5 minutes)

Chat prompt:

Load data/penguins.csv with readr. Show: how many rows, 
column names, and any missing values per column.

Expected output:

library(tidyverse)
penguins <- read_csv("data/penguins.csv")
cat("Rows:", nrow(penguins), "\n")
names(penguins)
penguins |> summarise(across(everything(), ~ sum(is.na(.))))

Step 2: Clean Data (5 minutes)

Chat prompt:

Remove rows with missing values. Show how many rows we had 
before and after.

Expected output:

penguins_clean <- penguins |> drop_na()
cat("Before:", nrow(penguins), "| After:", nrow(penguins_clean), "\n")

Step 3: Summary Statistics (5 minutes)

Chat prompt:

Group by species and show: count, average bill length, 
average body mass. Round to 1 decimal place.

Expected output:

penguins_clean |>
  group_by(species) |>
  summarise(
    n = n(),
    avg_bill = round(mean(bill_length_mm), 1),
    avg_mass = round(mean(body_mass_g), 1)
  )

Step 4: One Visualization (10 minutes)

Chat prompt:

Scatter plot: bill length (x-axis) vs flipper length (y-axis).
Color by species. Add title and axis labels. Use minimal theme.

Expected output:

penguins_clean |>
  ggplot(aes(x = bill_length_mm, y = flipper_length_mm, color = species)) +
  geom_point(alpha = 0.7, size = 2) +
  labs(
    title = "Penguin Measurements",
    x = "Bill Length (mm)",
    y = "Flipper Length (mm)"
  ) +
  theme_minimal()

If Something Goes Wrong

Error in your code?

Copy the full error message
Open Chat (Cmd+L)
Paste: I'm getting this error: [paste error]. Here's my code: [paste code]. What does it mean?

Plot doesn’t look right?

Select the plot code
Press Cmd+K
Type: Make this plot [describe what you want]

Next Steps

Pick one to extend your analysis:

Add a second plot (box plot of body mass by species)
Calculate correlation between bill length and flipper length
Compare just two species instead of all three
Save your results to a CSV file

Key Takeaways

Start simple — load, clean, summarize, plot
Use Chat for guidance — don’t try to memorize R syntax
Test after each step — catch mistakes early
Iterate — your first version won’t be perfect, and that’s OK

Resources

Previous: 2. Data Analysis & Visualization

Congratulations! You now know how to use AI tools to do real data analysis.