Categorical, Numerical, and Ordinal Variables

• CAT: Grade like A, B, C, D
• NUM: Numerical: SCORE: like 89.64
• ORD: Ordinal ordered categorical: D < C < B < A

Which plot to use

• It all depends on the variables
  • NUM x NUM = Scatter
  • CAT x CAT = Mosaic
  • CAT x NUM = Box
  • NUM = Bar plot, Histogram
  • CAT = Bargraph

What makes data interesting?

• Contradictory to our expectations? So called “Bayesian Prior”
• Outliers
• High Correlation
• What are Top K, Bottom K values?
• Look what I found - can’t wait to show you…
  • Salaries do not depend on education?
  • Salaries clearly are postively correlated with education
  • If groom and bride are born under the same sign THEN marriage has much higher chance to survive

Interesting vs Actionable?

Interesting:

• Wines from Montenegro are much more expensive than French Wines
• Californian wines are rated the highest
• Sweden has the highest cost of living
• Greatest basketball players are more than 6’7” or taller

Actionable:

• Honda has the best repair record
• Vegetarians live 3 years longer
• Lincoln tunnel traffic is higher than Holland tunnel traffic on weekends

Why look for patterns, trends?

• Actionable
  • We can do something based on the analysis which will benefit someone
• Data Cleaning
  • Biased data collection, errors, missing data
• Curiosity
  • Did you know that

How much R do I need to know?

• One Liners

• student_performance <- read.csv("MOODY.csv")

• boxplot(student_performance$SCORE, main='My first Boxplot')

• mosaicplot(moody$GRADE~moody$ON_SMARTPHONE)