Miscellaneous Graphs
Gokul Bhandari
Saturday, February 07, 2015
Correlation Plot
mcor <- cor(mtcars)
library(corrplot)
corrplot(mcor)
Correlation Plot- Customized
corrplot(mcor, method="shade", shade.col=NA, tl.col="black", tl.srt=45)
Generating a lighter palette
col = colorRampPalette(c("#BB4444", "#EE9988", "#FFFFFF", "#77AADD", "#4477AA"))
corrplot(mcor, method="shade", shade.col=NA, tl.col="black", tl.srt=45,
col=col(200), addCoef.col="black", addcolorlabel="no", order="AOE")## Warning in text.default(pos.xlabel[, 1], pos.xlabel[, 2], newcolnames, srt
## = tl.srt, : "addcolorlabel" is not a graphical parameter## Warning in text.default(pos.ylabel[, 1], pos.ylabel[, 2], newrownames, col
## = tl.col, : "addcolorlabel" is not a graphical parameter## Warning in title(title, ...): "addcolorlabel" is not a graphical parameter
# The data frame is only used for setting the rangePlotting a function
library(ggplot2)
p = ggplot(data.frame(x=c(-3,3)), aes(x=x))
p + stat_function(fun = dnorm)
Plotting a t-distribution
p + stat_function(fun=dt, args=list(df=2))
Defining a function
myfun <- function(xvar) {
1/(1 + exp(-xvar + 10))
}
ggplot(data.frame(x=c(0, 20)), aes(x=x)) + stat_function(fun=myfun)
shading a subregion under a curve
# Return dnorm(x) for 0 < x < 2, and NA for all other x
dnorm_limit <- function(x) {
y <- dnorm(x)
y[x < 0 | x > 2] <- NA
return(y)
}
# ggplot() with dummy data
p <- ggplot(data.frame(x=c(-3, 3)), aes(x=x))
p + stat_function(fun=dnorm_limit, geom="area", fill="blue", alpha=0.2) +
stat_function(fun=dnorm)
Defining a useful function
limitRange <- function(fun, min, max) {
function(x) {
y <- fun(x)
y[x < min | x > max] <- NA
return(y)
}
}
# This returns a function
dlimit <- limitRange(dnorm, 0, 2)
# Now we'll try out the new function -- it only returns values for inputs
# between 0 and 2
dlimit(-2:4)## [1] NA NA 0.39894228 0.24197072 0.05399097 NA
## [7] NAp + stat_function(fun = dnorm) +
stat_function(fun = limitRange(dnorm, 0, 2),
geom="area", fill="blue", alpha=0.2)
Creating a network graph
# May need to install first, with install.packages("igraph")
library(igraph)
# Specify edges for a directed graph
gd <- graph(c(1,2, 2,3, 2,4, 1,4, 5,5, 3,6))
plot(gd)
# For an undirected graph
gu <- graph(c(1,2, 2,3, 2,4, 1,4, 5,5, 3,6), directed=FALSE)
# No labels
plot(gu, vertex.label=NA)
Making the plot predictable
set.seed(229)
plot(gu)
Directed Graph with Fruchterman-Reingold Algorith
library(gcookbook) # For the data set
head(madmen2)## Name1 Name2
## 1 Abe Drexler Peggy Olson
## 2 Allison Don Draper
## 3 Arthur Case Betty Draper
## 4 Bellhop in Baltimore Sal Romano
## 5 Bethany Van Nuys Don Draper
## 6 Betty Draper Don Draper# Create a graph object from the data set
g <- graph.data.frame(madmen2, directed=TRUE)
# Remove unnecessary margins
par(mar=c(0,0,0,0))
plot(g, layout=layout.fruchterman.reingold, vertex.size=8, edge.arrow.size=0.5,
vertex.label=NA)
Undirected Graph- Circular Layout
g <- graph.data.frame(madmen, directed=FALSE)
par(mar=c(0,0,0,0)) # Remove unnecessary margins
plot(g, layout=layout.circle, vertex.size=8, vertex.label=NA)
Using Text Labels in a Network Graph
library(igraph)
library(gcookbook) # For the data set
# Copy madmen and drop every other row
m <- madmen[1:nrow(madmen) %% 2 == 1, ]
g <- graph.data.frame(m, directed=FALSE)
# Print out the names of each vertex
V(g)$name## [1] "Betty Draper" "Don Draper" "Harry Crane"
## [4] "Joan Holloway" "Lane Pryce" "Peggy Olson"
## [7] "Pete Campbell" "Roger Sterling" "Sal Romano"
## [10] "Henry Francis" "Allison" "Candace"
## [13] "Faye Miller" "Megan Calvet" "Rachel Menken"
## [16] "Suzanne Farrell" "Hildy" "Franklin"
## [19] "Rebecca Pryce" "Abe Drexler" "Duck Phillips"
## [22] "Playtex bra model" "Ida Blankenship" "Mirabelle Ames"
## [25] "Vicky" "Kitty Romano"plot(g, layout=layout.fruchterman.reingold,
vertex.size = 4, # Smaller nodes
vertex.label = V(g)$name, # Set the labels
vertex.label.cex = 0.8, # Slightly smaller font
vertex.label.dist = 0.4, # Offset the labels
vertex.label.color = "black")
Graph with Colored Edge
# This is equivalent to the preceding code
V(g)$size <- 4
V(g)$label <- V(g)$name
V(g)$label.cex <- 0.8
V(g)$label.dist <- 0.4
V(g)$label.color <- "black"
# Set a property of the entire graph
g$layout <- layout.fruchterman.reingold
plot(g)
# View the edges
E(g)## Edge sequence:
##
## [1] Henry Francis -- Betty Draper
## [2] Allison -- Don Draper
## [3] Don Draper -- Betty Draper
## [4] Candace -- Don Draper
## [5] Faye Miller -- Don Draper
## [6] Megan Calvet -- Don Draper
## [7] Rachel Menken -- Don Draper
## [8] Suzanne Farrell -- Don Draper
## [9] Hildy -- Harry Crane
## [10] Franklin -- Joan Holloway
## [11] Roger Sterling -- Joan Holloway
## [12] Rebecca Pryce -- Lane Pryce
## [13] Abe Drexler -- Peggy Olson
## [14] Duck Phillips -- Peggy Olson
## [15] Pete Campbell -- Peggy Olson
## [16] Playtex bra model -- Pete Campbell
## [17] Ida Blankenship -- Roger Sterling
## [18] Mirabelle Ames -- Roger Sterling
## [19] Vicky -- Roger Sterling
## [20] Kitty Romano -- Sal Romano# Set some of the labels to "M"
E(g)[c(2,11,19)]$label <- "M"
# Set color of all to grey, and then color a few red
E(g)$color <- "grey70"
E(g)[c(2,11,19)]$color <- "red"
plot(g)
Creating a Heat Map
head(presidents,3)## [1] NA 87 82str(presidents)## Time-Series [1:120] from 1945 to 1975: NA 87 82 75 63 50 43 32 35 60 ...#Convert it into data.frame
pres_rating <- data.frame(
rating = as.numeric(presidents),
year = as.numeric(floor(time(presidents))),
quarter = as.numeric(cycle(presidents))
)
head(pres_rating,3)## rating year quarter
## 1 NA 1945 1
## 2 87 1945 2
## 3 82 1945 3# Base plot
p <- ggplot(pres_rating, aes(x=year, y=quarter, fill=rating))
# Using geom_tile()
p + geom_tile()
# Using geom_raster() # looks the same, but a little more efficient
p + geom_raster()
#Customizing the heat map
p + geom_tile() +
scale_x_continuous(breaks = seq(1940, 1976, by = 4)) +
scale_y_reverse() +
scale_fill_gradient2(midpoint=50, mid="grey70", limits=c(0,100))