R Basics
Emily Mathew, 118348001, Michelle Nguyen, 116666107
1 Basic R manipulation.
- In the code segment below, assign the value 2 to variable x, and the value 6 to variable y. Then compute the sum ‘x+y’:
x <- 2
y <- 6
print(x+y)## [1] 8- Construct a list of all even numbers between 1 and 100, assign this to the variable ‘even’. Construct a list of 100 numbers, starting with 1, such that the difference between consecutive numbers is three, assign this to the variable ‘threes’. Construct a list of 100 equally spaced numbers between 0 and 2. Assign this to the variable x.
even <- seq(2,100,2)
threes <- seq(1, 100, 3)
x <- seq(0, 2, length.out = 100)- Define a function that computes the square of the input value, call this function square:
square <- function(num) {
num*num
}- Plot trigonometric functions sine and cosine for values of x between 0 and 10. Also plot the ‘square’ function for input values between 0 and 2.
x <- seq(0, 10, 0.1)
plot(x, sin(x), type="l", col="red")
lines(x, cos(x), col="blue")
val <- seq(0, 2, 0.1)
plot(val, square(val), type="l")
2 Working with data in R.
The dataset ‘iris’ is part of the in-built data repository of R. Use the ‘data’ command to load the iris dataset.
data(iris)The standard way of storing data in R is a two-dimensional dataframe. The columns correspond to the the variables and the rows correspond to observations. To get a sense what the data frame looks like, we use the ‘head’ command. Apply the head command to the dataframe iris.
head(iris)## Sepal.Length Sepal.Width Petal.Length Petal.Width Species
## 1 5.1 3.5 1.4 0.2 setosa
## 2 4.9 3.0 1.4 0.2 setosa
## 3 4.7 3.2 1.3 0.2 setosa
## 4 4.6 3.1 1.5 0.2 setosa
## 5 5.0 3.6 1.4 0.2 setosa
## 6 5.4 3.9 1.7 0.4 setosaWhat are the variable/column names in this dataframe? The variable/column names are Sepal.Length, Sepal.Width, Petal.Length, Petal.Width, and Species.
There is an easy way to find all the column names of a dataframe, using the ‘colnames’ command. Now, apply that command to the iris dataset.
colnames(iris)## [1] "Sepal.Length" "Sepal.Width" "Petal.Length" "Petal.Width" "Species"We will now work with the data. The first step is to be able to isolate a column/variable. We do that using the ‘$’ operator. Assign to x, the values of the column “Petal.Length”.
x <- iris$"Petal.Length"Note that x is now a list of numbers, which corresponds to the petal lenghts of the three different species of the flower iris.
Let us compute the summarry statistics for x, using the min, max, first, second, and third quartile, and the mean.
min(x)## [1] 1mean(x)## [1] 3.758quantile(x, .25)## 25%
## 1.6median(x)## [1] 4.35quantile(x, .75)## 75%
## 5.1max(x)## [1] 6.9There is an easy way to summarise these for the entire dataframe using the summary command. Use this command and compare the values that your manually computed above to the output of the ‘summary’ command.
summary(x)## Min. 1st Qu. Median Mean 3rd Qu. Max.
## 1.000 1.600 4.350 3.758 5.100 6.900We want to see if there is any relationship between the variables. To do so, it is important that we separate data based on the species. Use the filter command (from the library ‘dplyr’) to create three new dataframes callled “setosa”, “versicolor”, and “virginica”. Use the “head” command to view the first five rows of these data frames.
library(dplyr)##
## Attaching package: 'dplyr'## The following objects are masked from 'package:stats':
##
## filter, lag## The following objects are masked from 'package:base':
##
## intersect, setdiff, setequal, unionsetosa <- (filter(iris, Species == "setosa"))
versicolor <- (filter(iris, Species == "versicolor"))
virginica <- (filter(iris, Species == "virginica"))
head(setosa)## Sepal.Length Sepal.Width Petal.Length Petal.Width Species
## 1 5.1 3.5 1.4 0.2 setosa
## 2 4.9 3.0 1.4 0.2 setosa
## 3 4.7 3.2 1.3 0.2 setosa
## 4 4.6 3.1 1.5 0.2 setosa
## 5 5.0 3.6 1.4 0.2 setosa
## 6 5.4 3.9 1.7 0.4 setosahead(versicolor)## Sepal.Length Sepal.Width Petal.Length Petal.Width Species
## 1 7.0 3.2 4.7 1.4 versicolor
## 2 6.4 3.2 4.5 1.5 versicolor
## 3 6.9 3.1 4.9 1.5 versicolor
## 4 5.5 2.3 4.0 1.3 versicolor
## 5 6.5 2.8 4.6 1.5 versicolor
## 6 5.7 2.8 4.5 1.3 versicolorhead(virginica)## Sepal.Length Sepal.Width Petal.Length Petal.Width Species
## 1 6.3 3.3 6.0 2.5 virginica
## 2 5.8 2.7 5.1 1.9 virginica
## 3 7.1 3.0 5.9 2.1 virginica
## 4 6.3 2.9 5.6 1.8 virginica
## 5 6.5 3.0 5.8 2.2 virginica
## 6 7.6 3.0 6.6 2.1 virginica3 Boxplots for Iris Data
Use the boxplot function to show side-by-side boxplots for the variable “Petal.Lenght” across the three different species.
boxplot(Petal.Length ~ Species, data=iris)
Use the boxplot function to show side-by-side boxplots for the variable “Sepal.Width” across the three different species.
boxplot(Sepal.Width ~ Species, data=iris)
Note at least two observations that you can make looking at these boxplots. - Of all three flower species, the setosa has the smallest petal lengths, but the largest sepal widths. - The interquartile range for the petal length of a setosa is small, meaning most setosas have about the same size petals.
4 Data Visualization: Histograms
Consider the following sample data:
samp_data <- c(1.54,0.56,1.67,1.54,1.30,1.15,1.59,-0.22,1.01,1.03,1.02,1.77,1.13,0.15,1.02,
0.27,1.40,0.98,1.12,1.12,0.83,0.79,1.29,0.35,1.24,1.35,0.82,1.63,1.34,0.96,
1.37,0.14,1.43,2.00,1.13,1.39,1.18,0.99,0.65,0.88,1.11,1.63,1.53,0.91,0.02,
0.84,1.54,1.21,1.56,1.02,0.87,0.63,1.68,0.71,1.43,0.42,0.80,1.40,1.45,1.34,
1.57,0.98,0.99,1.04,1.15,0.47,0.82,0.89,0.78,1.02,1.56,1.67,1.32,1.23,0.46,
-0.06,0.90,0.94,1.60,0.52,1.04,1.79,0.66,0.76,1.47,0.68,1.38,0.35,1.30,1.30,
0.89,1.62,0.66,0.97,1.27,1.03,1.12,0.48,0.97,0.77,3.00,3.30,2.73,3.21,3.92,
3.33,2.25,2.58,3.45,2.86,3.01,3.27,2.65,3.48,3.18,2.01,3.42,2.13,3.15,2.77,
2.98,3.18,3.20,2.96,2.79,3.15,3.42,2.16,3.41,3.31,3.52,3.28,3.90,3.25,2.91,
2.98,3.25,3.02,2.52,2.97,2.31,3.30,2.50,3.19,2.46,2.82,2.83,2.48,2.55,3.18,
3.56,3.69,3.04,2.91,2.94,2.81,2.93,3.11,2.28,2.63,3.37,3.49,3.08,3.49,2.49,
2.89,3.15,3.21,2.79,3.46,3.56,3.33,3.15,2.18,3.08,2.24,4.31,3.23,3.35,3.18,
3.08,3.79,2.44,3.04,2.36,2.23,3.55,2.70,3.01,3.48,3.68,3.37,3.01,3.18,2.13,
3.39,3.37,2.86,3.34,3.16,2.54,3.58,3.63,2.43,3.94,2.80,2.41,2.25,2.87,3.17,
2.07,2.63,2.99,3.21,2.57,2.93,3.32,2.49,1.89,2.98,2.83,3.80,2.92,2.90,3.00,
3.10,2.41,2.66,2.77,2.69,2.99,2.50,3.21,2.98,3.45,3.83,3.00,2.54,3.86,2.80,
2.83,2.97,2.36,2.88,3.05,2.89,1.71,3.64,2.49,3.45,3.41,3.01,2.57,3.87,3.23,
2.57,2.56,2.85,2.83,3.19,3.15,3.17,2.90,3.29,3.46,2.43,3.04,2.81,3.28,2.82,
3.00,3.78,3.16,3.96,4.03,3.57,3.95,3.20,2.41,2.58,2.79,3.23,2.77,2.97,3.34,
3.13,3.78,2.70,3.14,3.25,2.26,3.99,3.10,3.76,2.92,2.69,3.58,3.11,3.59,2.74,
3.82,2.07,2.81,1.74,3.45,3.38,3.74,2.79,2.71,2.83,2.74,1.58,2.94,3.53,3.01,
2.21,2.61,2.95,2.58,3.92,2.68,2.85,3.06,2.86,3.43,3.16,3.02,2.29,3.74,3.49,
4.04,1.66,2.98,1.70,3.80,3.47,3.26,2.55,2.92,3.46,3.28,3.31,2.32,2.78,4.08,
3.78,2.49,3.26,2.88,3.23,3.15,2.62,3.73,3.03,2.78,3.00,2.61,2.48,2.03,3.04,
3.64,3.43,3.27,2.41,3.60,2.75,2.94,3.23,3.22,2.81,2.72,3.62,2.94,3.17,3.11,
2.60,3.11,3.45,2.37,3.07,3.09,3.90,2.86,3.23,3.50,2.84,2.17,2.58,3.57,2.95,
2.56,2.42,3.38,3.28,3.19,2.77,3.27,4.01,3.65,3.39)Use the ‘hist’ function to visualize samp_data. Furthermore, use different values of for the “breaks” parameter, which species the number of bins, to explore the shape of the data.
hist(samp_data)
hist(samp_data, breaks = 10)
hist(samp_data, breaks = 50)
hist(samp_data, breaks = 400)
hist(samp_data, breaks = 100)
What are some good values for the “breaks” parameter? Why? Since there are 400 values in samp_data, I experimented with bigger values to get a good idea of the distribution of data. Some good values for the breaks parameter are 50 and 100 because they show a detailed distribution of the data without being hard to understand or look at. When using 10, it was not that detailed and did not give extensive results like 50 and 100. When using 400, it was hard to look at all the measurements and understand the distribution of data.
Is the sample mean a good measure of location for this dataset? Why? Since this dataset is skewed to the left, the mean is not a good measure of location. From this dataset, we can assume an accurate measure of location would be around 3, since there is such a high frequency of 3s in the sample. However, the mean accounts for the frequencies of 0, 1, and 2, which causes the mean to be much smaller than we’d expect. As a result, the median would be a better measure of location for this dataset.