Live Code Execution with typeRun

Introduction

typeRun() is an enhanced version of typeR() that not only simulates typing animation but also executes your R code in real-time. This makes it perfect for:

🎓 Live teaching and workshops
🎤 Conference presentations
📹 Recording tutorials
🧪 Interactive demonstrations

Basic Usage

Simple Script Execution

The most basic usage is to run an R script with typing animation and live execution:

# Create a simple script
cat("# Data Analysis Demo
x <- 1:10
mean(x)
sum(x)
", file = "demo.R")

# Type and execute it
typeRun("demo.R")

What you’ll see:

Code types out character by character
Each line executes as it’s typed
Results appear immediately after each command

typeR::typeRun demo

Comparison: typeR vs typeRun

# typeR: Just displays, doesn't execute
typeR("demo.R")
# Output: Shows the code typing out (no results)

# typeRun: Types AND executes
typeRun("demo.R")
# Output: Shows code typing + execution results

Customizing the Experience

Control Typing Speed

# Slower typing for dramatic effect
typeRun("demo.R", delay = 0.1, jitter = 0.02)

# Faster typing for quick demos
typeRun("demo.R", delay = 0.02, jitter = 0.005)

Parameters:

delay: Base time between characters (seconds)
jitter: Random variation for natural feel

Limit Output Length

For long outputs, control how much is displayed:

cat("
# Long vector
long_vec <- 1:1000
long_vec

# Large data frame
big_df <- mtcars[rep(1:32, 10), ]
big_df
", file = "long_output.R")

# Show only first 5 elements/rows
typeRun("long_output.R", max_print = 5)

Working with Models

Linear Models

cat("
# Linear regression
model <- lm(mpg ~ hp + wt, data = mtcars)
summary(model)
", file = "model_demo.R")

typeRun("model_demo.R")

What happens:

model <- lm(...) executes silently (no output)
summary(model) displays the full summary with coefficients, R-squared, etc.

GLM and Other Models

typeRun() handles all model types intelligently:

cat("
# Logistic regression
glm_model <- glm(am ~ hp + wt, data = mtcars, family = binomial)
summary(glm_model)

# Poisson regression
pois_model <- glm(carb ~ mpg, data = mtcars, family = poisson)
summary(pois_model)
", file = "glm_demo.R")

typeRun("glm_demo.R")

Statistical Tests

The typeRun() function properly displays statistical test results using their formatted print methods:

library(typeR)

# Create a test script
cat('x <- rnorm(50, mean = 20, sd = 3)
tt <- t.test(x, alternative = "two.sided", conf.level = 0.95)
tt
', file = "test_ttest.R")

# Run it
typeRun("test_ttest.R", delay = 0.01)

This will display the t-test results in their proper formatted output:

    One Sample t-test

data:  x
t = 45.185, df = 49, p-value < 2.2e-16
alternative hypothesis: true mean is not equal to 0
95 percent confidence interval:
 18.68753 20.42711
sample estimates:
mean of x 
 19.55732

The same applies to other statistical objects like anova(), aov(), lm(), etc.

Interactive Control

Pause and Resume

During execution, you have full control:

Press ESC (or Ctrl+C) to pause
You’ll see: Enter choice (1 resume or 2 stop):
Enter 1 to continue from where you paused
Enter 2 to stop completely

typeRun("long_script.R")
# Press ESC while running
# Enter 1 to resume
# Or enter 2 to stop

This is perfect for:

Answering questions during presentations
Explaining specific code sections
Debugging during demonstrations

Working with R Markdown and Quarto

R Markdown Files

typeRun() intelligently handles .Rmd files:

# Create an R Markdown example
typeRun("report.Rmd")

Behavior:

✅ Types and shows all text
✅ Executes only R code chunks
✅ Skips YAML headers
✅ Preserves narrative flow

Quarto Documents

Works identically with .qmd files:

typeRun("analysis.qmd")

Advanced Features

Custom Evaluation Environment

Run code in an isolated environment to keep your workspace clean:

# Create custom environment
my_env <- new.env()

# Run in that environment
typeRun("demo.R", envir = my_env)

# Check what was created
ls(my_env)
ls()  # Your global environment is unchanged

Suppressing Library Messages

Package loading messages are automatically suppressed:

cat("
library(ggplot2)  # No startup message shown
library(dplyr)    # Clean output

# But code works normally
mtcars %>% head()
", file = "packages.R")

typeRun("packages.R")
# Shows only the actual results, not package messages

Output Handling Details

What Executes Silently

These don’t produce output (cleaner demos):

cat("
# Assignments - silent
x <- 1:10
y <- mean(x)

# Plots - execute but don't clutter console
plot(x, y)
hist(x)

# Package loading - no startup messages
library(stats)
", file = "silent.R")

typeRun("silent.R")

What Shows Output

cat("
# Direct values
1:10          # Shows the vector

# Function calls
mean(1:10)    # Shows: [1] 5.5

# Print statements
print('Hello') # Shows: [1] 'Hello'

# Model summaries
summary(lm(mpg ~ hp, data = mtcars))  # Shows full summary
", file = "visible.R")

typeRun("visible.R")

Practical Examples

Teaching Linear Regression

cat("
# Load data
data(mtcars)
head(mtcars, 3)

# Visualize relationship
plot(mtcars$hp, mtcars$mpg,
     xlab = 'Horsepower',
     ylab = 'Miles per Gallon',
     main = 'MPG vs Horsepower')

# Fit model
model <- lm(mpg ~ hp, data = mtcars)
summary(model)

# Add regression line
abline(model, col = 'red', lwd = 2)

# Predictions
new_data <- data.frame(hp = c(100, 150, 200))
predict(model, new_data)
", file = "teaching_demo.R")

typeRun("teaching_demo.R", delay = 0.08)

Data Analysis Workflow

cat("
# 1. Load and explore
data <- iris
str(data)

# 2. Summary statistics
summary(data)

# 3. Visualization
boxplot(Sepal.Length ~ Species, data = data,
        main = 'Sepal Length by Species',
        col = c('lightblue', 'lightgreen', 'pink'))

# 4. Statistical test
aov_result <- aov(Sepal.Length ~ Species, data = data)
summary(aov_result)

# 5. Post-hoc test
TukeyHSD(aov_result)
", file = "analysis_demo.R")

typeRun("analysis_demo.R", delay = 0.06, max_print = 8)

GLM Demonstration

cat("
# Binary outcome: Manual transmission (am)
# Predictors: HP and weight

# Fit logistic regression
logit_model <- glm(am ~ hp + wt, 
                   data = mtcars, 
                   family = binomial(link = 'logit'))

# Model summary
summary(logit_model)

# Odds ratios
exp(coef(logit_model))

# Predicted probabilities
mtcars$pred_prob <- predict(logit_model, type = 'response')
head(mtcars[, c('am', 'hp', 'wt', 'pred_prob')])
", file = "glm_example.R")

typeRun("glm_example.R", max_print = 6)

Tips and Best Practices

1. Prepare Your Script

Before presenting:

# Test your script normally first
source("demo.R")

# Then test with typeRun
typeRun("demo.R")

2. Timing Considerations

# For 5-minute demos: ~0.05 delay
typeRun("demo.R", delay = 0.05)

# For dramatic reveals: ~0.1 delay
typeRun("demo.R", delay = 0.1)

# For quick reviews: ~0.02 delay
typeRun("demo.R", delay = 0.02)

3. Output Management

# Lots of output? Limit it
typeRun("demo.R", max_print = 5)

# Models in script? They'll show nicely
# Plots in script? They'll render silently

4. Keep Scripts Focused

# Good: One concept per script
typeRun("01_data_loading.R")
typeRun("02_visualization.R")
typeRun("03_modeling.R")

# Less good: One giant script
typeRun("everything.R")  # Too long, hard to control

Troubleshooting

Script Not Found

# Error: The file does not exist
typeRun("missing.R")

# Solution: Check file path
file.exists("demo.R")
list.files(pattern = "\\.R$")

Unexpected Output

# If output looks wrong, test the script normally first
source("demo.R")

# Then try with typeRun
typeRun("demo.R")

Want to Stop Mid-Execution

# Press ESC, then enter 2
# Or press Ctrl+C (on some systems)

Comparison Table

Feature	`typeR()`	`typeRun()`
Typing animation	✅	✅
Code execution	❌	✅
Shows output	❌	✅
Interactive pause/resume	❌	✅
Output truncation	❌	✅
Custom environment	❌	✅
Model summary handling	N/A	✅
Library message suppression	N/A	✅