EDA: Exploratory Data Analysis

• Variation is the tendency of measured values of a variable to change measurement-to-measurement.

Building a Visualization (some missing point)

• legend:

  • legend("the location in the graph, like bottomright", legend, fill: the color of different level, cex)
    • pch: 点形状
    • lty: 线形状

• add lines:

  • abline(int, slope)
  • lines( )
  • example: add line separately

  cuts        <- levels(small_diam$cut)
  col_counter <- 1 
  for (i in cuts) {
  	this_cut  <- small_diam$cut == i
    this_data <- small_diam[this_cut, ]
    this_lm   <- lm(log(this_data$price) ~ log(this_data$carat))
  
    abline(this_lm, col = col_counter)
    col_counter <- col_counter + 1 
  }
  
  

• point:

  • points(x, y)

• Example: simulate 10k standard normal

# simulate 10000 standard normal

Z <- rnorm(10000)
head(Z)

# plot a hist of Z w/ the true density overlayed on the plot

hist(Z, breaks = 30)

# create a normal curve
## need to consider the y-axis is FREQUENCY or DENSITY

z.plot <- seq(-5, 5, length = 10, probability = T)
plot(z.plot, dnorm(z.plot), col = "purple", lty = 1, lwd = 1)

# plot sin(x) cos(2x)+1 from -2pi,2pi in same graph
x.plot <- seq(-2*pi, 2*pi, length = 100)
plot(x.plot, sin(x.plot), col = "green", type = "l", ylim = c(-1.5, 2.5))
lines(x.plot, cos(x.plot*2)+1, col = "red", type = "l")

Wk4: Visualizations Cont. & ggplot