Published
Edited
Dec 7, 2021
countries = FileAttachment("custom.geo-3.json").json()
northAmerica = ({
type: "FeatureCollection",
features: countries.features
.filter((d) => d.properties.continent == "North America")
.filter((d) => d.properties.name != "United States")
})
states = FileAttachment("gz_2010_us_040_00_5m.json").json()
//for converting long-lat to pixels
projection = d3.geoAlbers().center([-10, 50]).scale(400)
//center on the Americas
//.rotate([90, -15, 0])
height = 600
elSalvador = FileAttachment("El Salvador.csv").csv()
mexico = FileAttachment("Mexico.csv").csv()
guatemala = FileAttachment("Guatemala.csv").csv()
//Create an empty array, and combine (concatenate) all migration data together into it
migrationData = [].concat(elSalvador, mexico, guatemala)
//Find unique source countries from concatenated array
supportedSources = Array.from(new Set(migrationData.map((d) => d.Hometown)))
supportedGeoData = supportedSources.map((d) =>
countries.features.find((e) => e.properties.name == d)
)
//Find all different column headers from first row in CSV file, and then remove the two that aren't time-related
supportedYears = Object.keys(migrationData[0])
.filter((d) => d != "Hometown")
.filter((d) => d != "Destination")
statesWithMigrationData = ({
type: "FeatureCollection",
features: states.features
.map((state) => {
return {
...state,
migration: migrationData.filter(
(d) => d.Destination == state.properties.NAME
)
};
})
.map((state) => ({
...state,
centroid: d3.geoCentroid(state),
trails: supportedGeoData.map((country) => ({
type: "Feature",
properties: {
source: country.properties.name,
target: state.properties.NAME,
migration: migrationData
.filter((d) => d.Destination == state.properties.NAME)
.filter((d) => d.Hometown == country.properties.name)[0],
distance: d3.geoDistance(
d3.geoCentroid(state),
d3.geoCentroid(country)
)
},
geometry: {
type: "LineString",
coordinates: [d3.geoCentroid(country), d3.geoCentroid(state)]
}
}))
}))
})
parseInt(
statesWithMigrationData.features[0].migration.filter(
(d) => d.Hometown == "Mexico"
)[0]["2010"]
)
migrationScale = d3
.scaleLinear()
.domain(
d3.extent(
migrationData.filter((d) => selectedSources.includes(d.Hometown)),
(d) => parseInt(d[selectedPeriod])
)
)
.range([0, 1])
//show min and maxes to ensure everything is working
d3.extent(
migrationData.filter((d) => selectedSources.includes(d.Hometown)),
(d) => parseInt(d[selectedPeriod])
)
//Create some buttons for each source country
viewof selectedSources = Inputs.checkbox(supportedSources, {
value: ["El Salvador", "Mexico", "Guatemala"]
})
//Create some buttons for each column from CSV file
viewof selectedPeriod = Inputs.radio(supportedYears, { value: "2010" })
//To read about what's happening here...
//https://observablehq.com/@observablehq/introduction-to-generators
{
while (animated == true) {
for (const year of supportedYears) {
viewof selectedPeriod.value = year;
viewof selectedPeriod.dispatchEvent(new Event("input"));
yield Promises.delay(5000);
}
}
}
sourceScale = d3.scalePoint().domain(supportedSources).range([0, 1])
targetScale = d3
.scalePoint()
.domain(states.features.map((d) => d.properties.NAME))
.range([0, 1])
activeMigrationValues = migrationData
.filter((d) => selectedSources.includes(d.Hometown))
.map((d) => d[selectedPeriod])
.map((d) => parseInt(d))
//for how many seconds the animaton will take, based on population
durationScale = d3.scaleLinear().domain(activeMigrationValues).range([20, 10])
{
//svg area
const svg = d3.create("svg").attr("width", width).attr("height", height);
svg
.append("rect")
.attr("width", width)
.attr("height", height)
.attr("fill", "#555b6e");
//rotated group to hold countries
const g = svg.append("g");
// .attr("transform", "rotate(-90," + width / 2 + "," + height / 2 + ")");
//the function that takes geographic data and converts to SVG paths
const path = d3.geoPath().projection(projection);
//draw countries
g.selectAll(".countries")
.data(northAmerica.features)
.enter()
.append("path")
.attr("d", path)
.attr("stroke", (d) =>
supportedSources.includes(d.properties.name) ? "white" : "#bee3db"
)
.attr("fill", (d) =>
selectedSources.includes(d.properties.name)
? d3.interpolatePlasma(sourceScale(d.properties.name))
: "#89b0ae"
)
.attr("stroke-width", 1);
//draw countries
g.selectAll(".states")
.data(statesWithMigrationData.features)
.enter()
.append("path")
.attr("d", path)
.attr("stroke", "#bee3db")
.attr("stroke-width", 1)
.attr("fill", "#89b0ae");
//walk through states
for (let state of statesWithMigrationData.features) {
const stateArcs = state.trails
.filter((d) => selectedSources.includes(d.properties.source))
.filter((d) => d.properties.migration);
//draw arcs
g.selectAll(".arcs")
.data(stateArcs)
.enter()
.append("path")
.attr("d", (d) => path(d))
.attr("fill", "none")
.attr("stroke-width", 2)
.attr("stroke", (d) =>
d3.interpolatePlasma(sourceScale(d.properties.source))
)
.style("stroke-dasharray", "2 20")
.style("stroke-dashoffset", (d) => Math.random() * 100)
.style(
"animation",
(d) =>
"dash " +
durationScale(parseInt(d.properties.migration[selectedPeriod])) +
"s linear infinite"
);
}
return svg.node();
}
html`
<style>
@keyframes dash {
to {
stroke-dashoffset: -100;
}
}
</style>`
One platform to build and deploy the best data apps
Experiment and prototype by building visualizations in live JavaScript notebooks. Collaborate with your team and decide which concepts to build out.
Use Observable Framework to build data apps locally. Use data loaders to build in any language or library, including Python, SQL, and R.
Seamlessly deploy to Observable. Test before you ship, use automatic deploy-on-commit, and ensure your projects are always up-to-date.