This assignment consists of six parts:
Create a script called hw02-4.R and save it in your hw02 folder.
After you complete each exercise, commit and push your R script to your remote repo. See Part 0 for instructions. Do not push this document.
In R, a factor is vector data used for categorical variables, also called discrete data. Discrete data can belong only to one of a limited number of categories, such as plant, animal, fungus, or bacteria. The Linnean classification hierarchy used by biologists is another example of categorical data. A species can belong to only one genus, one family, one order, and so on. Factors are different from continuous data that can have an unlimited number of possible values. R handles factors and continuous variables differently as you will learn during this course.
To create factors in R, you first create a vector with the names of your discrete categories. You then pass that vector to the factor() function. Study this example.
# Genotype can be homozygous or heterozyous. This vector contains
# the genotype of six individuals.
genotype_vector <- c("Heterozygote", "Heterozygote", "Homozygote", "Heterozygote", "Homozygote", "Heterozygote")
genotype_factor <- factor(genotype_vector)
genotype_factor## [1] Heterozygote Heterozygote Homozygote Heterozygote Homozygote
## [6] Heterozygote
## Levels: Heterozygote HomozygoteThe output lists the elements of the vector and also Levels. Levels are the unique categories that factor() found in the vector.
For this exercise, you will assume that you counted the number and species of bumble bees (genus Bombus) that visitied a single flowering plant in one hour. Your notes show the order of the visitors:
Add code to your script to
Make a character vector called bee_visitors_vector with these seven visitors. Your vector must reflect the visiting order listed above.
Use factor() to make a factor from bee_visitors_vector. Assign the new factor to bee_visitors_factor
Print out bee_visitors_factor to be sure you have the order and three levels. A incorrect number of levels suggests you might have spelled a name incorrectly.
# Make a character vector called `bee_visitors_vector`
# Make a `bee_visitors_factor` factor from `bee_visitors_vector`.
# Print out bee_visitors_factor
#Your results should look like this. Notice that the levels are arranged alphabetically. R does this automatically, which may not be what you want. You’ll learn later how to arrange factors in the order you want.
## [1] kirbiellus kirbiellus flavifrons kirbiellus bifarius flavifrons kirbiellus
## Levels: bifarius flavifrons kirbiellusCategorical variables can be ordinal or nominal. Ordinal categories are ordered. For example, the bee species listed above have proboscis lengths that can be categorized as short, intermediate, or long An order is implied because short < intermediate < long lengths. You can choose to order the categories from short to long or long to short. What’s important is that an order is implied.
Nominal categories (nomen = name) do not have an implied order. The three species of bees are nominal categories. There is no implied order to the names of the three species. Bombus kirbiellus is not any more or less a bumble bee than B. flavifrons or B. bifarius. They’re just different.
Study the example below. The Bacillus factor is nominal because no order is implied for the bacteria. factor() defaults to alphabetical order for the levels. The allozyme factor is ordinal because allozymes move through an electrophoretic gel at different speeds (here, slow, medium, fast). Notice that the order of the levels is shown as Slow < Medium < Fast.
# Bacillus bacteria
bacillus_vector <- c("subtilis", "cereus", "pasteurii", "sphaericus")
bacillus_factor <- factor(bacillus_vector)
bacillus_factor## [1] subtilis cereus pasteurii sphaericus
## Levels: cereus pasteurii sphaericus subtilis# Allozymes are slow, medium, or fast
allozyme_vector <- c("Medium", "Slow", "Fast", "Slow", "Medium")
allozyme_factor <- factor(allozyme_vector, order = TRUE, levels = c("Slow", "Medium", "Fast"))
allozyme_factor## [1] Medium Slow Fast Slow Medium
## Levels: Slow < Medium < FastPay attention to the arguments for factor() in the allozyme example. In addition to the vector, the order = TRUE argument tells the function that order matters. The levels = c("Slow", "Medium", "Fast") argument tells the function the specific order of the levels.
allozyme_vector <- c("Medium", "Slow", "Fast", "Slow", "Medium")
allozyme_factor <- factor(allozyme_vector, order = TRUE, levels = c("Slow", "Medium", "Fast"))How would you rewrite the call to factor() to arrange the allozyme levels from fast to slow?
Here are the proboscis lengths for the three species of Bombus you surveyed.
Add code to your script to
Create a vector of seven lengths that matches the order of the bees in your nominal vector above (the order you observed them). Name that vector proboscis_vector.
Create an ordinal factor called proboscis_factor, with the lengths ordered from long to short.
Display proboscis factor to be sure you have it correct.
Hint: Remember that case matters.
# Create `proboscis_vector` with seven elements
# Create an ordered `proboscis_factor` ordered from long to short.
# Display the contents of `proboscis_factor
## Create `proboscis_vector` with seven elements
proboscis_vector <- c("Long", "Long", "Intermediate", "Long", "Short", "Intermediate", "Long")
# Create an ordered `proboscis_factor` ordered from long to short.
proboscis_factor <- factor(proboscis_vector, ordered = TRUE, levels = c("Long", "Intermediate", "Short"))
# Display the contents of `proboscis_factor
proboscis_factor## [1] Long Long Intermediate Long Short
## [6] Intermediate Long
## Levels: Long < Intermediate < Short#Note: The default for factor() is to create unordered (nominal) factors. You must specify levels and pass the ordered = TRUE argument to the function to create ordinal factors.
A useful function that you will use often is summary() which provides a summary of the information contained in a variable. The example below shows how to apply summary() to allozyme_factor to summarize the number of observations of each level.
summary(allozyme_factor)## Slow Medium Fast
## 2 2 1Add code to your script to
summary() function to bee_visitors_factor and proboscis_factor. You do not have to save the results to a variable.Factors are simply categorical vectors. If you want to extract a particular element, you can use square brackets [] like you would with other vectors. Study these examples.
# Show all the elements for review
allozyme_factor
# Extract specific elements
allozyme_factor[4] # Get the fourth element
allozyme_factor[2:4] # Get elements 2 through 4
allozyme_factor[c(1,3,5)] # Get elements 1, 3, and 5
# Extract slow allozyme elements using a boolean vector
slow_allozymes <- allozyme_factor == "Slow"
allozyme_factor[slow_allozymes]You have to be careful using boolean operators like > and <= because they only work on ordered factors where some level is greater than or less than another level. Nominal factors are unordered so “greater than” and “less than” do not have any context. Study these example.
# Show bacillus (unordered) and allozyme (ordered) factors for review
bacillus_factor## [1] subtilis cereus pasteurii sphaericus
## Levels: cereus pasteurii sphaericus subtilisallozyme_factor## [1] Medium Slow Fast Slow Medium
## Levels: Slow < Medium < Fast# Extract allozymes that are faster than slow. This is just one method.
not_slow_allozymes <- allozyme_factor > "Slow"
allozyme_factor[not_slow_allozymes]## [1] Medium Fast Medium
## Levels: Slow < Medium < Fast# Test to see which is greater: medium or fast
az1 <- allozyme_factor[1] # medium allozyme
az3 <- allozyme_factor[3] # fast allozyme
# FALSE because allozyme level is Slow < Medium < Fast.
# Medium is less than Fast.
az1 > az3 ## [1] FALSE# The reverse is TRUE
az3 > az1 # Could also test as az1 < az3## [1] TRUE## Now, test with underordered bacillus factor
##
# This works because there is a level for "cereus"
bacillus_factor == "cereus"## [1] FALSE TRUE FALSE FALSE# This does not because one species is not greater than another
bacillus_factor > "cereus"## Warning in Ops.factor(bacillus_factor, "cereus"): '>' not meaningful for factors## [1] NA NA NA NAAdd code to your script to do the following.
Extract from bee_visitors_factor the elements that are not kirbiellus, using a boolean technique. If necessary, review 02_vectors to review different boolean operators. Try to use the shortest code possible.
Extract from bee_visitors_factor the second, and fourth through sixth elements. Use the : for the range. You do not have to save this result to a variable.
Test whether the first element of proboscis_factor is longer than the last element of proboscis_factor. Follow the example for allozyme_factor above. You do not have to save this result to a variable.
Optional challenge: Redo the previous test of the first and last elements but use length() to obtain the position of the last element in proboscis_factor. Remember you can look up help for functions using ?.
# Extract from `bee_visitors_factor` the elements that are *not* kirbiellus
# Extract from `bee_visitors_factor` the second, and fourth through sixth elements. Use the colon for the range.
# Test whether the first element of `proboscis_factor` is longer than the last element of `proboscis_factor`
# CHALLENGE
#Be sure you have used comments throughout your code to identify sections and to describe what you are doing, then commit and push your hw02-4.R script to GitHub.