Midterm Exam (2026)

Accept this assignment through the link in Canvas.

The exam is a comprehensive take-home exam.

Each year, Sandhill Cranes migrate through Nebraska on their way from their Winter areas in the southern US and Mexico to their Summer breeding grounds in Northern Canada, Alaska, and Siberia.

Hundreds of cranes roost in the shallows of the Platte river.

Closeup of five Sandhill cranes.

The photographs show Sandhill Cranes migrating through Nebraska.

Can we see this migration in hobbyists’ data? iNaturalist is a non-profit organization that provides an online platform for the users of their app to submit geo-tagged observations of plants and animals and share their findings with researchers and the public.

The file cranes.csv contains results from this platform using the search term “Sandhill Crane”.

While you should not need it for answering the questions here, a description of some of the variables and the data structure is available at the iNaturalist Open Data documentation.

For all of the following questions, provide answers and include your code. You can work in R or in python or in a mix of the two. Make sure that the final document you submit renders without errors. In the rendered document avoid any lengthy output.

1. Read the data (10 points)

Load the data into the object cranes using R or python.

1. Make sure that you are working with an object with 10536 rows and 39 columns.
2. How many numeric variables, how many text variables does the data set contain? Are there any other types of variables?

# code chunk for R

# python code chunk

2. First overview (20 points)

Answer the following questions and provide the code that allows you to answer the question.

1. Which variable has the highest number of missing values? What is the percentage of missing values across all variables?
2. How many different users (user_id) submitted observations? Which user submitted the highest number of observations?
3. The variable scientific_name contains the scientific name of the observed species. Create a summary of the fifteen (sub-)species most frequently observed.

# code chunk for R

# python code chunk

3. Visual Inspection (25 points)

1. “Antigone canadensis” is the scientific name of the Sandhill Crane. How many different scientific names in scientific_name contain the substring Antigone canadensis? Create an object sandhill that contains only Sandhill Cranes (and their subspecies).
2. Create visualizations that contain the variables below in a meaningful way. Describe each of the visualizations you create and what you learn about the data they show:

• latitude, longitude contain the geographic latitude and longitude
• observed_on date on which the observation was made
• year, day year and day of the week an observation was made
• common_name common name of the observed animal

3. Nebraska fails between -104 and -95.3 degrees longitude and 40.0 and 43.0 degrees latitude. Create an object nebraska that contains all of the Sandhill Crane observations between these coordinates. How many observations are in nebraska? Do the observations follow I-80 or the Platte River more closely? Discuss your answer.

# code chunk for R

# python code chunk

4. How many days since January 1? (20 Points)

You are asked to write a function days_since (date) that calculates the number of days it has been since January 1st of that year.

Assume that the date is written in the format like “d2020-03-14”, i.e. year, month, and day are given by 4, 2, and 2 digits, respectively. In case the month or the year are single digits, a leading zero is added. The string starts with a ‘d’; the values for year, month, and day are separated by “-”.

Watch out for leap years! February has an extra day in 2020 and in 2024:

Year	Jan	Feb	Mar	Apr
2019	31	28	31	30
2020	31	29	31	30
2021	31	28	31	30
2022	31	28	31	30
2023	31	28	31	30
2024	31	29	31	30
2025	31	28	31	30

Test your function: the output from days_since("d2021-04-03") should be 93; days_since("d2024-03-27") should be 87; and days_since("d2025-03-14") should be 73.

1. Write a detailed list of steps to calculate from the string named date the number of days since Jan 1 of the same year.

2. Implement the function days_since in R or python.

3. Apply the function days_since to the variable observed_on_string in the cranes data, and find the average number of days since Jan 1 for each of the years.

# R chunk

# python chunk

5. A time line (25 points)

1. Ensure that the variable observed_on is a date variable. (If you work on R, use the lubridate package to convert the variable observed_on to a date object. If you work in python, use the datetime package.)
   Create a data set with the following data summaries:

• a variable n_obs that contains the number of observations for each day,
• a variable year_day that contains the number of days of that year since Jan 1 (in case you can’t get your function from 4 to work, use lubridate::yday or pandas .dt.dayofyear)
• a logical variable weekend that is TRUE if the day is a Saturday or a Sunday and FALSE for a week day.

2. Create a scatterplot of the number of observations mapped to y, the year_day variable mapped to x and weekend mapped to the color of the points.

3. Calculate the average number of observations that are made on weekends and on weekdays. Based on the summary statistics and the visual, discuss the statement: on average, there are significantly more observations made during weekends than on weekdays

# R chunk
library(lubridate)

# python chunk
import datetime