R源代码学习资料（全套）

# Up and Running with R # Ex03_01 # Creating bar charts for categorical variables # Use dataset "social_network.csv" which records the # gender and age of 202 online survey respondents # along with their preferred social networking site # and an estimate of how many times they log in # per week. # Create data frame "sn" from CSV file w/headers sn <- read.csv("social_network.csv", header = T) # R doesn't create bar charts directly from the categorical # variables; instead, we must first create a table that # has the frequencies for each level of the variable. # The "table" function is able to create this table from # the variable, which we specify as sn$Site. That is, we # first give the name of the data frame, then $, then the # name of the variable. (Following Google's style guide, # it is better to state explictly the data frame than to # use the shortcut function "attach," which can lead to # confusion.) site.freq <- table(sn$Site) # Creates table from Site barplot(site.freq) # Creates barplot in new window ? barplot # For more information on customizing graph # To put the bars in descending order, add "order": barplot(site.freq[order(site.freq, decreasing = T)]) # Draw the bars horizontally (but turn off decreasing) barplot(site.freq[order(site.freq)], horiz = T) # Make Facebook blue and all others gray by specifying a # vector with named colors for gray and RGB for Facebook blue # fbba = "Facebook blue/ascending" for horizontal bars # And now breaking code across lines for clarity fbba <- c(rep("gray", 5), rgb(59, 89, 152, maxColorValue = 255)) barplot(site.freq[order(site.freq)], horiz = T, col = fbba) # Turn off borders with "border = NA" # Add title with "main" with "\n" to break line # Add subtitle with "sub" barplot(site.freq[order(site.freq)], horiz = T, # Horizontal col = fbba, # Use colors "fbba" border = NA, # No borders xlim = c(0, 100), # Scale from 0-100 main = "Preferred Social Networking Site\nA Survey of 202 Users", xlab = "Number of Users") # Immense amount of control available through # "par {graphics}" # RStudio gives option of saving as image file in # several formats and it's easier to change sizes