Coding self-assessment

Harris Coding Camp

Summer 2023

As part of the statistics curriculum, you will be asked to analyze data using the programming language R.
R is an open source language that is widely used by data analysts and data scientists. In coding camp and
coding lab, we provide an introduction to R coding focused on data analysis.
This is a self-assessment. If you feel comfortable completing this assignment by yourself (with the help of
Google), then you are free to skip the coding camp and coding lab. Otherwise, you can use this to pick the
right track for you.

Task 1:1
1. Install R and RStudio.
2. Install the package readxl and tidyverse.
3. Adjust the following code block to read in the provided data set,
incarceration_counts_and_rates_by_type_over_time.xlsx

library(tidyverse)
library(readxl)
setwd(<Put path to file here>)
incarceration_data <- read_xlsx("incarceration_counts_and_rates_by_type_over_time.xlsx",
range = "A7:CO10") %>%
rename("type" = ...1) %>%
pivot_longer(`1925`:`2016`, names_to = "year", values_to = "counts")

4. What does the code library(readxl) do and why is it necessary?

5. Why do you need to set a working directory (setwd())?
6. How many vectors are there in the dataset? How many observations?

7. Briefly explain the difference between vectors, lists and data frame.

If you had trouble with readxl, we provide a csv file as well. You can load the data with the following code:

incarceration_data <- read_csv("incarceration_counts_and_rates_by_type_over_time.csv")

1 Copying and pasting from the pdf will create issues in syntax–particularly it messes up the type of quotes used. We provide

a file with this code in a text file. Alternatively, you can re-type the code or copy and paste and then fix syntax issues.

1
Task 2:
We want to analyze state prison counts by decade. We’ll prepare the data in the following ways. Store the
following changes in a new tibble (data frame) called state_data.

1. Add a column called decade that reflects which decade the observation comes from.
2. Filter the data so that you only have data from State prisons.
3. Use select to reorder the columns so that your data is organized as below:

## # A tibble: 10 x 4
## type counts decade year
## <chr> <dbl> <dbl> <dbl>
## 1 State prisons 85239 1920 1925
## 2 State prisons 91188 1920 1926
## 3 State prisons 101624 1920 1927
## 4 State prisons 108157 1920 1928
## 5 State prisons 107532 1920 1929
## 6 State prisons 117268 1930 1930
## 7 State prisons 124118 1930 1931
## 8 State prisons 125721 1930 1932
## 9 State prisons 125962 1930 1933
## 10 State prisons 126258 1930 1934

4. Finally, find out the mean, standard deviation, max and min value of counts for all observations from
State prisons.

Task 3:
In this section, you’ll use group_by() and summarize() to answer questions about state prison counts by
decade.

C −C
1. Which decade saw the largest percentage growth in State prisons? Measure percent growth as deCd ds
s
where Cde is the count at the end of decade and Cds is the start of the decade). You may consider
using the first() and last() functions.

## # A tibble: 10 x 2
## decade percentage_growth
## <dbl> <dbl>
## 1 1920 0.262
## 2 1930 0.365
## 3 1940 -0.0490
## 4 1950 0.245
## 5 1960 -0.0644
## 6 1970 0.581
## 7 1980 1.15
## 8 1990 0.725
## 9 2000 0.129
## 10 2010 -0.0553

2
Task 4:
You want to make a graph visualizing the change in incarceration counts in the United States over time.

incarceration_data %>%
ggplot(???) +
geom_???() +
labs(???)

Adjust the code above in order to reproduce the following graph, including the choice of both axes, labels
on both axes, choice of line type and title.
Incarceration counts (total population on a single day) over time

1e+06

type
counts

Federal prisons
Local jails
State prisons
5e+05

0e+00

1925 1950 1975 2000

year

Task 5:
Miscellaneous tasks – We leave the data behind and test skills.

1. Take numbers <- rep(seq(-9, 10, 1), 10). Show that the mean of the vector numbers is .5 and
the sum of the components of numbers is 100.
2. Combine the strings assigned to left and right into a single string using an R function.

left <- "Harris"

right <- "School of Public Policy"

3. Use ifelse() function to add a column called index to the incarceration_data and assign high if
count >= 300000 and low otherwise.

3
Task 6:
1. For loops: Take numbers <- rep(seq(-9, 10, 1), 10). Using a for-loop, save the square of each
number in a new vector called numbers_squared.
2. For loops: Take numbers. Using a for-loop, save the square of each number and add random noise
using a call to rnorm(1, sd = 5) in a new vector called noisy_numbers_squared.
You should be able to reproduce the graph below:

numbers_data <- tibble(numbers = numbers,

noisy_numbers_squared = noisy_numbers_squared)

numbers_data %>%
ggplot(aes(x = numbers, y = noisy_numbers_squared)) +
geom_point() +
geom_smooth()

## ‘geom_smooth()‘ using method = ’loess’ and formula = ’y ~ x’

80
noisy_numbers_squared

40

−5 0 5 10
numbers
3. Functions: Write a function called notice_gpa that takes gpa as an input and does the following:

• if gpa less than 2, prints: “Your GPA is gpa. You are on academic probation.”
• else if gpa is greater than or equal to 3.5, prints: “Your GPA is gpa. You made the Dean’s list.
Congrats!”
• otherwise, prints: “Your GPA is gpa”.

4
notice_gpa <- function(gpa) {
if (...) {
...
} else if (...) {
...
} else {
...
}
}

# When running each of the following, you should get different results!
notice_gpa(1.9)
notice_gpa(3.5)
notice_gpa(3)