At the encouragement of some friends, I've decided to try the 2023 30-Day Map Challenge. This repository contains the code and maps I've created during this challenge.
For my first map, I decided to plot the area where I grew up: Brevard County, Florida. This map depicts the locations of historic bridges in Brevard County. The data is sourced from the Florida Geographic Data Library (BAR Historic Bridges in Florida - Oct 2023).
The colors for this map were selected from the first image of the Sebastian Inlet Bridge in this article. The Sebastian Inlet Bridge is identified with a triangle on this map.
For day 2, I decided to plot the road system in Rome, Georgia. (Honestly, I just though of the saying "all roads lead to rome" and decided to run with it.) The spatial line file came from the Georgia GIS Clearinghouse.
For day 3, I decided to plot the housing vacancy rates across census tracts in Washington, D.C. I also overlaid the boundaries for D.C.'s eight wards and added identifying labels. The data came from the Washington D.C. Open Data portal.
I'm a huge fan of the Eurovision Song Contest, and this is the first but likely not the last ESC-related map I'm going to make for this challenge. There's some essential context needed here: Eurovision is open to countries with eligible television broadcasters in the European Broadcasting Area, a region which encompasses Europe as well as parts of North Africa and Western Asia. Over the years, various countries have participated in Eurovision which aren't necessarily in Europe, like Armenia, Azerbaijan, Georgia, Israel, and Morocco. It's this last country which is the focus of my bad map. Morocco only entered Eurovision once, in 1980, when they sent the song Bitaqat Hub (بطاقة Řب) (sometimes spelled Bitakat Hob) by Samira BensaĂŻd. Unfortunately, it did extremely poorly. In the Eurovision Song Contest, countries award points to each other, and are unable to award points to themselves. Morocco received only 7 points, all from Italy. (For comparison, the winning entry in 1980, What's Another Year by Johnny Logan representing Ireland, received 143 points). This resulted in Morocco finishing 18th of 19 countries, barely ahead of Finland. The poor showing so offended the Moroccan public broadcaster MRT that they decided to withdraw from the contest and to date have never sent another entry.
This map shows all the countries which have awarded points to Morocco in the Eurovision Song Contest. The two colors used (red for Morocco and green for Italy) are the two colors of the Moroccan flag. If you don't like this map, maybe you'll like Bitaqat Hub better. I suggest you listen to it; it's a fantastic fusion of disco and traditional Arabic music styles, and Samira BensaĂŻd is a stellar artist. Morocco may have only sent one Eurovision entry, but it's a good one.
The shapefile for this map came from opendatasoft.
[To be added if I get around to it]
In the years following World War II and the Korean War, South Korea was an extremely poor and underdeveloped nation. Government ministers, looking to Germany's economic development and growth as an example, sought to lead their own country in a similar direction. Germany had the "Miracle on the Rhine", and South Korea would have the "Miracle on the Han River", so named for the large river which flows through the nation's capitol and largest city, Seoul.
Many of the businesses which sprung up during this period of economic growth are known as chaebols. Chaebols are large companies owned by a single extended family, similar to the Rockefellers, Vanderbilts, or Pritzkers in the U.S. Several chaebols built their corporate headquarters or flagship stores along the Han River in Seoul. This map shows the locations of some of those headquarters.
The shapefile for the municipalities of Seoul comes from this GitHub repository of South Korean GIS data. The shapefile for the Han River originally comes from the Food and Agriculture Organization of the United Nations, but it's a large file, so the version included here in this repo is filtered to only the Han River. The list of chaebols and the families who own them came from the "List of major chaebols by family group" table on the Wikipedia page for chaebols.
The colors for this map were taken from the first image on the Wikipedia page for the Hibiscus syriacus, also known as the Rose of Sharon, which is the national flower of South Korea.
I have to be honest, I don't feel particularly proud of this map. This is a map of neighborhood clusters in northwest Washington D.C., colored according to the metro line of the stop closest to them. In northwest D.C., all of the stops on the blue line are also on the orange and silver lines. These are presented as blue for simplicity. The green line overlaps substantially with the yellow line, so a similar approach was taken. The colors of the map are the official WMATA hex codes. The shapefile for the neighborhood clusters can be found here. The data for the metro stops is taken from the WMATA API by way of a tutorial I wrote for the folium package in python during my master's degree. The data can be found here, as can the tutorial. At some point, I may try to rework this into something better, but for now I'm focusing on finishing as many maps as I can for this challenge.
Sure, Toto really had something with the whole "bless the rains down in Africa" thing, but what were those rains really like? For my Day 8 map, I chose to visualize the average rainfall in millimeters/year between 1983 and 2000. Why 1983? Because "Africa" by Toto came out in 1982. Why 2000? Because that's where the data ended.
This map clearly shows that the eastern parts of South Africa got a higher amount of annual rainfall on average than the western regions of the country. However, the data I found did not have records for several municipalities, particularly in the east, hence the greyed-out locations. I've identified the top 5 and bottom 5 municipalities for average annual rainfall and labeled them. Some of the weather reporting stations matched to the same municipality, which is why you might see two labels pointing to the same polygon but listing different values. They're capturing different parts of that polygon.
The colors for this map are drawn from the South African flag: red for the dry parts, green for the wet parts, black and blue for the fill scale, gold for the borders and background, and white for the header and footer panels. The data for this map came from the South African Environmental Observation Network (SAEON), as well as this random person's upload of municipality boundaries and Princeton University for the nation-level shapefile of South Africa I used in preparing the data for this map.
For this map, I chose to visualize DW-Nominate scores. DW-Nominate is a method political scientists use to quantify a legislator's position on a left-right spectrum relative to other legislators based on their votes in a given legislative session. In the U.S., scores below 0 indicate more left-leaning legislators, while scores above 0 indicate more right-leaning legislators (like a number line). This map uses a hexagon shapefile from the Daily Kos (available through this link) and DW-Nominate scores from voteview.com. While the version shown here in the README is static, this map is interactive, and I've included an HTML version in the Day 9 folder which has labels for each congressional district with the name and party of its representative as well as their DW-Nominate score. Despite the interactive nature of this map, this was honestly one of the easier ones I've done for this challenge so far.
For this map, I re-used some data I personally compiled for a project for my master's degree. That project centered on the social media presences of county public health departments across the U.S. during the COVID-19 pandemic. I was mainly looking if counties which had these accounts (and furthermore, which had active accounts) had notably different rates of COVID infection or death. (Spoiler alert: they didn't.) But the data lives on, even if it is only accurate as of December 2021. This is a filled polygon map of U.S. counties, colored by whether or not they have a Facebook account for their public health department, a Twitter account, both, or neither. Some really populous counties, like Bernalillo County, New Mexico, don't have social media accounts for their public health department, while some small rural counties, like the ones in Nebraska and Idaho, do have these accounts. This is because many rural counties opted to consolidate their public health departments into multi-county public health districts for efficiency, and these multi-county districts have social media accounts. If you look at the code, you can see which of my repos I'm pulling this data from, and that repo also has data on those multi-county departments as well as a shapefile of their boundaries.
Is this a bit of a cop-out? Yes. Do I care? No.
[To be added if I get around to it]
Don't get it twisted, I don't like Orinoco Flow as a song all that much. That being said, the title did provide the inspiration for my Day 12 map, which shows cities and towns in Colombia and Venezuela along the Orinoco River. Locations are identified by labels and sized according to population, and labels are colored by country.
Data for the river came from the UN Food and Agriculture Organization, and the coordinates for the different cities came from simplemaps.com.
Supreme Court Justice Louis Brandeis once famously referred to state legislatures as "laboratories of democracy", venues for the citizens of the several states to attempt new and interesting policy solutions in order to address the issues of the day. My choropleth map attempts to visualize this somewhat; here, states are shaded based on the average number of bills introduced by a legislator in 2021. This is calculated as Total # of Bills Introduced/(Size of Lower Chamber + Size of Upper Chamber).
Data on bill filing is taken from this FiscalNote report and data on state legislature size is taken from the Wikipedia article "Comparison of U.S. state and territory governments".
This was a really quick one. I reused a lot of my code from day 10 here (and may end up putting some of that into a function to make it easier to create inset maps like this going forward).
I said I'd make another Eurovision map, and I did. This map shows the different cities which have hosted the Eurovision Song Contest from its inception in 1956 (Lugano, Switzerland) to the upcoming 2024 contest (to be held in Malmö, Sweden). Points are colored based on the decade in which that city most recently hosted, and are sized according to the number of times they've hosted.I recommend viewing the interactive version in the folder; the city markers there have popups which name them and include the years they've hosted as well as links to the winning entries from those years. Some essential context for Eurovision: part of the prize for winning the contest is the right to host the following year. This has been a practice since 1958, and there have only been a few exceptions: 1972 (Edinburgh, United Kingdom), when Monaco was unable to host as the incumbent winner, 1974 (Brighton, United Kingdom), after Luxembourg was unable to host a second time in a row following their victories in 1972 and 1973, 1980 (The Hague, The Netherlands) after Israel was unable to host a second time following their victories in 1978 and 1979, and 2023 (Liverpool, United Kingdom), when Ukraine was unable to host after their 2022 victory due to the ongoing war.
The link for 2020 (Rotterdam, the Netherlands) is empty, since that contest was cancelled due to the COVID-19 pandemic. The link for the 2024 contest in Malmö is also empty since that contest has not happened yet. The link for the 1969 contest (Madrid, Spain) goes to a playlist, since the 1969 contest ended in an unprecedented four-way tie between France, Spain, the Netherlands, and the United Kingdom. The 1990 contest was hosted in Zagreb in what is now Croatia; at the time, however, this was the first and only contest hosted in the former Yugoslavia.
If you've never been exposed to Eurovision before, I recommend checking out the 1974 winner (Brighton, United Kingdom) or the 1988 winner (Dublin, Ireland). Those are probably the two most internationally-famous artists. Some of my personal favorite winning songs are 1993 (Millstreet, Ireland), 2007 (Helsinki, Finland), and my single favorite Eurovision song of all time, 2016 (Stockholm, Sweden). Happy listening!
For this map I plotted a data set of museum locations in Washington, D.C. alongside WMATA metro stations using the ggspatial package, which allows for easy use of OpenStreetMap tiles to create static plots. This is a relatively simple and unsophisticated map, but it does the job.
The data for the WMATA stations is the same data used in my Day 7 map, and the data for the museum locations comes from opendata.dc.gov.
This map has two facets. The left side shows which areas of New Zealand farm the most sheep, cast as hexagonal polygons. The right side shows the 2017 season for the Canterbury Rams, a basketball team which plays in the New Zealand National Basketball League; specifically, it shows the average number of points scored against the opposing teams at different stadiums across New Zealand (calculated as mean(Rams Score/Opponent Score)).
I was hoping to see that playing in parts of the country with lots of sheep meant the Rams would perform better, perhaps out of a sense of interspecies/mascot-driven solidarity. This does not appear to be the case. The southernmost part of the southern island, which has more sheep than anywhere else in New Zealand, saw one of the Rams' weaker average performances, while their best average score (achieved at the Trafalgar Centre) took place in a part of the country where there are few if any sheep. Suffice it to say, the Rams do not appear to derive strength on the court from the local presence of their namesake animals.
The data for the sheep counts came from data.mfe.govt.nz, and the data for the Canterbury Rams 2017 season came from nznbl.basketbal.
One of the oddest details about the Eurovision Contest is that there's no requirement that a country's song be written by people from that country. As such, there tends to be a decent amount of cross-border songwriting (though that's decreased in recent years). This map shows the countries which participated in the 2023 contest, filled by whether or not their songs were written entirely by lyricists and composers from that country. The blue arrows indicate a relationship; for example, one of the songwriters for this year's Irish entry is Swedish, so there's an arrow from Sweden to Ireland. Fans of the contest might notice that Serbia is shaded yellow, even though the lyricist and composer for that song, Luke Black, is Serbian. This is because "national origin" for the purposes of this data set means the country where a songwriter lives and works, and not where they're originally from. Luke Black is based in the United Kingdom, hence the arrow.
The data for this map came from this tweet by Twitter user @EMursiya. I transcribed it into a spreadsheet and am including it in the folder for this map (note: "birth country" refers to the countries indicated by flags on @EMurisya's map, while "country origin" refers to the countries listed in parentheses in the annotations on that map).
[To be added if I get around to it]
In fairness, this map took me a little longer than five minutes, but not much longer. It's a map of the United States at the county level, and counties are colored based on a regex pattern match for color names (i.e. Greenwood County is colored green, Spink county is colored pink, Iredell county is colored red, and so on). I used the tigris package to import the shapefile, and since it comes with the county names built in, no additional data was needed here.
I return once again to the place I grew up, Brevard County, Florida. This time, I've chosen to visualize the topography of the county. Granted, Florida as a whole is very flat, so the range of elevations (5 meters to 80 meters) doesn't span a particularly large range of values, but there's still enough to make for a decent map. Merritt Island (the land mass off the coast) is very low-lying, while some of the more inland parts of the county are more elevated.
The data for this map came from the Florida Department of Environmental Protection Geospatial Open Data Portal.
Growing up in Florida, one of the things I learned in school was the importance of mangrove trees in the local ecosystem. These trees are able to live in salt water, and use their roots to filter out the salt from the water. The red mangrove species which is particularly common in Florida has large roots which form habitats for local fish and marine life. (Here's a picture, courtesy of the Brevard Zoo,)
I chose to visualize mangrove coverage in south Florida as of 2020 for my raster map. The data comes from the [Advanced Land Observing Satellite Research and Application Project ALOS]. I merged and cropped this data to create the raster in the day21_raster folder, in order to save memory and keep the file size manageable.
When we look at maps of the natural world, many of the rivers depicted flow north-to-south, like the Mississippi River, the Volga River, or the Ganges River. The Nile River in northeast Africa, however, flows south-to-north, with tributaries starting in southern Uganda and west Ethiopia, joining in Sudan, and flowing northward through Egypt. This map presents the Nile, but oriented how we expect to see rivers; that it, it has been inverted, with its sources in the north and the delta in the south. Tributaries and sections of the river are colored by name (though the "Northern Nile" label is not officially used, and I applied it here mainly to delineate that section of the river basin from the others). The blue line running north-to-south is the main body of the river.
The data for the main body of the Nile River came from naturalearth.com, as did the country shapefiles. The data for the Nile tributaries came from the United Nations Food and Agricultural Organization, like the other river-based maps I've made.
Time is a dimension, right? That's what I'm going with for this map: longitude, latitude, and time, Three dimensions, like the challenge says. This gif shows the movement of the mean center of population in the United States from the first decennial census in 1790 to the most recent one (as of this map) in 2020. The states also become shaded darker as their population increases. I wanted to include a legend, but for some reason gganimate made it jump around a lot, and the effect was kind of nauseating. It's bad enough that the title moves the way it does. I also wanted the transitions to be smoother, but attempting to use any of the other transition types in gganimate basically stalled out the render process, so transition_manual was the best option. The data for the coordinates for the mean center of population come from the Wikipedia article linked above, and the data for state population comes from this Wikipedia article with tables of states and their populations across different census years.
This map shows the locations of historically black colleges and universities in the United States. HBCUs, as opposed to majority-black institutions, must have been founded before 1964. Many of these colleges and universities can trace their origins back to the mid-19th century or early 20th century, when they were established by or named for either free Black Americans (like William White, who cofounded Morehouse College in Atlanta) or anti-segregationist white Americans (like Laura Spelman Rockefeller, who financially supported Spelman College in Atlanta in its early years), or in some cases as land grant universities, like Bowie State University in Maryland or Langston University in Oklahoma. Some HBCUs, like Bluefield State University in West Virginia, are now majority-white. Regardless, these institutions are still considered HBCUs because historically their primary purpose was to serve and educate Black students.
The data for this map comes from the "Current Institutions" table on the Wikipedia page for the List of Historically Black Colleges and Universities. The version of this map included above is static, but there's an interactive version in the folder which lists the names and founding dates of each HBCU.
[To be added if I get around to it]
There's a series of children's books about a woman named Amelia Bedelia, who variously works as a housekeeper or nanny, or does other odd jobs. Amelia Bedelia always interprets figures of speech literally; when she's asked to draw the curtains, she takes out her sketchbook and some colored pencils. I've taken a bit of an Amelia Bedelia approach to this particular map. I know that the "minimal" theme means a map with very few features, but I've instead made a map of a very small (one might say "minimal") place: Monaco. Monaco is the smallest sovereign nation on earth besides the Vatican City. It occupies an area less than 1 square mile on the southern coast of France, and has fewer than 40,000 residents, of whom fewer than 10,000 are actually Monegasque (a delightful demonym, I might add). Despite being so small, Monaco has its own language (Monegasque), though its official language, as used in nearly all government business and day-to-day life, is French.
This map shows the nine wards of Monaco, labeled by name and shaded according to their population (with red as the gradient in honor of the Monegasque flag). The background color is a shade known as Monaco Blue. I've also labeled the Monte Carlo Casino, Monaco's largest and most famous tourist attraction and cultural center, with a small diamond.
The data for this map came from the Humanitarian Data Exchange.
This map displays census tracts and wards in D.C. Each tract is filled with dots colored according to generation, from Generation Alpha to the Silent Generation. (Ranges are approximate, and counts were calculated based on available data). Each dot represents 50 people. I was hoping this map would reveal some interesting geographic patterns. It looks like D.C. is filled with older people, particularly baby boomers and members of the silent generation. There appear to be some concentrations of Gen Xers in parts of northwest DC, while millenials look to be clustered in the eastern parts of the district.
The data for this map came from the American Community Survey by way of the D.C. Open Data Portal.
First and foremost, I want to give credit to Kyle Walker (@walkerke). I had some ideas for this map, but I wasn't quite sure how to execute them, or even if they'd produce a result I could feel good about. Then, I saw Kyle's Day 28 map on Linkedin, where he combines a map of the U.S. with a bar chart using the ggiraph package to create an interactive visualization, and I took inspiration from that. If I had more time, I'd love to tinker with the formatting some more on this one, but I think it does an alright job of conveying the data. I wanted the map to mirror the bar graph, so visualizing average temperature seemed like a natural choice. States with lower average temperatures are toward the top of the bar chart (Alaska excepted), and as temperature increases, latitude decraeses. This is reflected in the color of the bars and the fill for the states. The version in the README is static, but there is an interactive version in the Day 28 folder (it's probably easier to read, though it might run a little choppy).
The data for this map came from the National Centers for Environmental Information, the National Weather Service, and world-weather.info.
These maps show the prevalence of Americans in same-gender relationships across the United States and Puerto Rico. The numbers are taken from the 2020 decennial census. The first map shows raw counts per states, while the second one shows the percentage of each state's population in a same-gender relationship. It should be noted that the counts/percentages are derived from two variables in the census: DP1_0116C (count of Americans in same-gender marriages) and DP1_0118C (count of Americans in same-gender partnerships). For each map, I've labeled the top five and bottom five states.
The data for these maps comes from the 2020 U.S. Census.