ENVS363/563.3 - A Computational Essay 2023/24

ENVS363/563.3 - A Computational Essay 2023/24

Overview and Instructions

Due Date: 8th January 2024

50% of the final mark

Overview

Here’s the premise. You will take the role of a real-world GIS analyst or spatial data scientist tasked to explore datasets on the San Francisco Bay Area (often just called the  Bay Area) and find useful insights for a variety of city decision-makers. It does not matter if you have never been to the Bay Area. In fact, this will help you focus on what you can learn about the city through the data, without the influence of prior knowledge. Furthermore, the assessment will not be marked based on how much you know about the San Francisco Bay Area but instead about how much you can show you have learned through analysing data. You will need contextualise your project by highlighting the opportunities  and  limitations  of  ‘old’  and  ‘new’  forms  of  spatial  data  and  reference  relevant literature.

Format

A computational essay using Quarto. The assignment should be carried out fully in Quarto.

What is a Computational Essay?

A computational essay is an essay whose narrative is supported by code and computational results that  are  included  in  the  essay  itself.  This  piece   of  assessment   is equivalent to  4,000  words. However, this is the overall weight. Since you will need to create not only narrative but also code and figures, here are the requirements:

•     Maximum of 1,000 words (ordinary text) (references do not contribute to the word

count). You should answer the specified questions within the narrative. The questions should be included within a wider analysis.

•     Up to five maps or figures (a figure may include more than one map and will only count as one but needs to be integrated in the same overall output)

•     Up to one table

There are three kinds of elements in a computational essay.

1.     Ordinary text (in English)

2.    Computer input (R code)

3.    Computer output

These three elements all work together to express what’s being communicated.

Submission

You must submit 1 electronic copy of your assessment via Canvas by the published

deadline. The format of the file must bean html document. Please do not include your name anywhere in the documents.

•    Please refer to the ENVS363/563 Assessment criteria. This document includes the parts you should include in your Computational Essay.

Data

The assignment relies on datasets and has two parts. Each dataset is explained with more detail below.

•    Data made available on Murray Cox’s website as part of his “Inside Airbnb” project which you can download (http://insideairbnb.com/). The website periodically publishes

snapshots of Airbnb listings around the world. You should Download the San Francisco data, the San Mateo data and the Oakland data. These are all part of the Bay Area.

Please Note: that for best results you will need to drop some of the outliers.

•   Socio-economic variables for the Bay Area. Source: American Community Survey (ACS) 2016-2020, US Census Bureau. Observations: 1039; Variables: 472; Years: 2016-2020.

o A subset of variables from the latest ACS has already been retrieved for you in ACS_2016_2020_vars.csv. However, you have access to ALL variables in the

American Community Survey (ACS) 2016-2020 through the R package Tidycensus.

o You are strongly recommended to use the census API in the R package

Tidycensus to extract your variables of interest instead of the csv.  For more information about the ACS (2016-2020) you can have a look at:

https://www.census.gov/data/developers/data-sets/acs-5year.htmland https://api.census.gov/data/2020/acs/acs5/variables.html .

If you want to visualise some aspects at different Subnational Administrative boundaries, you can download USA boundaries fromGADM. You can also find other geodata for the Bay Area  in the Berkeley Library.

IMPORTANT - Students of ENVS563 will need to source, at least, two additional datasets relating to San Francisco or the Bay Area. You can use any dataset that will help you complete the tasks    below but, if you need some inspiration, have a look at the following:

•    Geodata for the Bay Area in theBerkeley Library.

•    San Francisco Open Data Portal:https://datasf.org/opendata/

•    Data World:https://data.world/datasets/san-francisco

•    NASA Data:https://earthdata.nasa.gov/earth-observation-data/near-real-time/hazards- and-disasters/air-quality

Part 1 - Common

1.1   Collecting and importing the data

1.1.1        Import and explore

1.2 Preparing the data

1.2.1       What CRS are you going to use? Justify your answer.

1.3   Discussion of the data

•    Present and describe the datasets used for this project.

1.4 Mapping and Data visualisation

1.4.1       Airbnb in the BAY AREA at Neighbourhood Level

•     Summarise the data. Using Bay Area zipcodes/ ZCTAs obtained fromBerkeley Library. This is slightly different from the Airbnb neighbourhood file. Obtain a count of listings by neighbourhood.

• Map 1.1: Number of listings per zipcode. Explore the spatial distribution of the data using choropleths. Style the layers using a colour ramp.

• Map 1.2: Average price per zipcode. Explore the spatial distribution of the data using choropleths. Style the layers using a colour ramp.

Justify your data classification methods and visualization choices. You should include these maps in  your   assessment  submission.  The   maps  should  be  well-presented  and  include   a   short description.

Questions to answer within your analysis: How does the Inside Airbnb data compare to other ‘new’ forms  of  spatial  data?  Discuss  the  potential  insights  and  biases,  as  well  as  opportunities  and  limitations of the Airbnb data.

1.4.2. Socio-economic variables from the ACS data

Select two variables from American Community Survey data. These could be but are not limited to population density, median income, median age, unemployed, percentage of black population, percentage of Hispanic population or education level. See the Appendix in this document for help. If you chose to calculate  population percentages, make sure you standardise the table  by the population size of each tract.

• Map2: Explore the spatial distribution of your chosen variables using choropleths. Style the variables using a colour ramp. Justify your data classification methods and visualization choices. You should include these maps in your assessment submission. The maps should be well-presented and include a short description.

Questions to answer within your analysis. Comment on the details of your map and analyse the results. What are the main types of neighbourhoods you identify? Which characteristics help you delineate this typology? What can you say about the spatial distribution of your socio-economic   variable of interest? If you had to use this classification to evaluate where Airbnbs would cluster, what would your hypothesis be? Why?

For some stylised (not necessarily accurate) facts about the Bay Area seehere.

1.4.3. Combining Datasets

• Map 3: Plot the natural logarithm of price (ln of price) of Airbnbs in the San Francisco Bay Area together (point plot) with one of your chosen socio-economic variables of interest    at zipcode level using ggplot ortmap or mapsf (polygon plot). There are various ways of  doing this. The maps should be well-presented.

Questions to answer within your analysis. Comment on the details of your map and analyse the results. Does this map tell you more about the relationship between Airbnb location/price and your socio-economic variable of choice? Explain your answer.

1.4.4. Autocorrelation

• Map 4: Explore the degree of spatial autocorrelation. Describe the concepts behind your approach and interpret your results.

Part 2 - Chose your own analysis

For this one, you need to pick one of the following three options. Only one, and make the most of it.

Please Note: This part of the assignment can be done on the Bay Area as a whole or you can zoom in on one of the counties. For example, you could just focus on San Francisco.

1.     Create a geodemographic classification and interpret the results. In the process, answer the following questions:

•    What are the main types of neighbourhoods you identify?

•    Which characteristics help you delineate this typology?

•    If you had to use this classification to target areas in most need, how would you use it? why?

2.    Create a regionalisation and interpret the results. In the process, answer at least the following questions:

•    How is the city partitioned by your data?

•    What do you learn about the geography of the city from the regionalisation?

•    What would one useful application of this regionalisation in the context of urban policy?

3. Use the OpenStreetMap package toosmdata download Point of Interest (POIs) Data for the Bay Area or San Francisco. Using this this data, complete the following tasks:

•    Visualise the dataset appropriately and discuss why you have taken your specific approach

•    Use DBSCAN to identify areas of the city with high density of  POIs, which we will call areas of interest (AOI). In completing this, answer the following questions:

o What parameters have you used to run DBSCAN? Why?

o What do the clusters help you learn about areas of interest in the city?

o Name one example of how these AOIs can be of use for the city. You can take the perspective of an urban planner, a policy maker, an operational

practitioner (e.g. police, trash collection), an urban entrepreneur, or any

other role you envision.

Resources to help you. See also suggested bibliography in slides throughout the course.

• https://www.r-bloggers.com/2017/11/programming-meh-lets-teach-how-to-write-

computational-essays-instead/

• https://rmarkdown.rstudio.com/

• https://www.rstudio.com/wp-content/uploads/2015/02/rmarkdown-cheatsheet.pdf

• https://vizual-statistix.tumblr.com/post/114850050736/i-find-the-spread-of-airbnb-to-be- as-fascinating

• https://carto.com/blog/airbnb-impact/

• https://cran.r-project.org/web/packages/biscale/vignettes/biscale.html

Appendix

American Community Survey (ACS) 2016-2020, US Census Bureau. Observations: 1039; Variables: 472; Years: 2016-2020