28:null 29:["$","title",null,{"children":"Should we fear fluoride? | Grant M | Observable"}] 2e:T530,### Introduction Fluoride is a chemical widely used for the purposes of dental hygiene. It is mainly found in toothpaste and other dental products, but it is possible to consume it via other natural sources as well. One such natural source is in water, and people consume a notable amount of fluoride by this method. Not all fluoride found in water is natural, however, as the fluoridation of community water supplies is a common practice. The American Dental Association highly recommends this practice and has linked it directly to a reduction in tooth decay by around 25% (Antoine, 2024). The intake of fluoride, in large quantities, however can directly cause skeletal fluorosis (Cleveland 2022). Additionally, there have been strong claims correlating it to a reduction in IQ for early childhood development (Taylor 2025). As a result, many states have been debating over banning the process of putting fluoridating their water supplies. This includes the state of Utah, which had a ban recently pass in their HB81 bill. Yet, there are pretty generous limits to the amount of fluoride that can be in water. It also seems unlikely that we are over saturating our water supply enough to warrant a ban on what seems to be a process with an overall positive impact on our society. Thus, this leads to the following hypothesis.2f:T93b,const data = ydatabase[yearmap[select_year]]; const isLarge = d => +d["ADJUSTED_TOTAL_POPULATION_SERVED"] > 1000000; const small = data.filter(d => !isLarge(d)); const large = data.filter(isLarge); const yearly_database = Plot.plot({ width: 975, height: 610, projection: "albers-usa", color: { type: "linear", domain: [0, 1.25, 1.5, 2.5, 4], range: ["#2b6fd6", "#f2d43d", "#e0301e", "#8b0000"], clamp: true, legend: false, label: "Fluoride (mg/L)", ticks: [0, 0.7, 1.5, 3, 4], }, marks: [ Plot.geo(statesGeoByMonth, {fill: "#eee", stroke: "#fff"}), Plot.dot(small, { x: "INTPTLONG", y: "INTPTLAT", fill: "VALUE MG/L", fillOpacity: 0.75, r: 3.5, // tip:true }), Plot.dot(large, { x: "INTPTLONG", y: "INTPTLAT", fill: "VALUE MG/L", fillOpacity: 1, r: {value: d => +d["ADJUSTED_TOTAL_POPULATION_SERVED"] *0.000004, scale: null}, stroke: "#363737", // tip:true, strokeWidth: 2.5 }), Plot.tip( data.filter(d => +d["ADJUSTED_TOTAL_POPULATION_SERVED"] >= 100), Plot.pointer({ x: "INTPTLONG", y: "INTPTLAT", maxRadius: 20, title: d => `${d["SYSTEM_NAME"]} ${d["STATE_CODE"]} \nFluoride (mg/L): ${ (+d["VALUE MG/L"]).toFixed(2)} ${+d["VALUE MG/L"] <= 1 ? "🟢" : +d["VALUE MG/L"] <= 1.5 ? "🟡" : "🔴"} \nPopulation Served: ${+d["ADJUSTED_TOTAL_POPULATION_SERVED"]}` }) ), Plot.tip( data.filter(d => +d["ADJUSTED_TOTAL_POPULATION_SERVED"] >= 1000000), Plot.pointer({ x: "INTPTLONG", y: "INTPTLAT", maxRadius: 5, title: d => `${d["SYSTEM_NAME"]} ${d["STATE_CODE"]} \nFluoride (mg/L): ${ (+d["VALUE MG/L"]).toFixed(2)} ${+d["VALUE MG/L"] <= 1 ? "🟢" : +d["VALUE MG/L"] <= 1.5 ? "🟡" : "🔴"} \nPopulation Served: ${+d["ADJUSTED_TOTAL_POPULATION_SERVED"]} \nSystem Type: ${d["SYSTEM_TYPE"]}` }) ), ] }); display(html`

${yearly_database.legend("color", { label: "Fluoride (mg/L)", ticks: [0, 0.7, 1.5, 2, 3, 4], tickFormat: d => d >= 4 ? "4+" : d.toString() })}

0.7 mg/L: US recommended level
1.5 mg/L: WHO guideline limit
4.0 mg/L: Dangerous to consume for long periods

`); display(yearly_database); 30:T526,const plot = Plot.plot({ width: 975, height: 610, projection: "albers-usa", color: { type: "linear", scheme: "turbo", clamp: true }, title: "Mean Fluoride versus Population", subtitle: `${countf} communities who serve over ${population_limit} people with more than ${fluoride_limit} mg/L of fluoride. Approximately ${ppl_sum.toFixed(0)} total people served or ${((ppl_sum.toFixed(0)/3429000000)*100).toFixed(2)}% of the US population.`, marks: [ Plot.geo(statesGeo, {fill: "#eee", stroke: "#fff"}), Plot.dot(means_min.filter(d => d.meanFluoride >= fluoride_limit).filter(d => d.meanPopulation >= population_limit), { x:"lon", y:"lat", fill: "black", }), Plot.density(means_min.filter(d => d.meanFluoride >= fluoride_limit).filter(d => d.meanPopulation >= population_limit), { x:"lon", y:"lat", fill: "red", fillOpacity: 0.4, thresholds: [0.19], }), Plot.tip( means_min.filter(d => d.meanFluoride >= fluoride_limit).filter(d => d.meanPopulation >= population_limit), Plot.pointer({ x: "lon", y: "lat", maxRadius: 5, title: d => `${d["name"]} \nFluoride (mg/L): ${ (+d["meanFluoride"]).toFixed(2)} \nPopulation Served: ${(+d["meanPopulation"]).toFixed(2)}` }) ), ] }); display(plot); 31:T4ad,I believe the data gather does not support the hypothesis. It stands to reason that, while strict, a 0.7 mg/L limit on fluoride in the water affects a very substantial part of the population (children). Additionally, filtering for small communities (below around 5000) does not strongly affect this outcome. Simply put, consuming 0.7 mg/L of fluoride regularly does not appear to be healthy and is affecting 30% of the population. Therefore, it stands to reason that we should take some greater action in restricting the amount of fluoride in the water supply. While it can still be a powerful tool, it seems like smaller communities manage their fluoride in a more efficient way. It also seems hypofluoridation is concentrated to areas in the midwest specifically. This is not to degrade the benefits of fluoridation, as this study in no way refutes the substantial amount of evidence for that claim. Rather, it is to show that, if we believe fluoride is harmful in certain quantities, it is likely producing a negative impact in our society. However, it should also be noted for users convinced of stricter limits like 1.5 mg/L, that the hypothesis is completely supported under this assumption.32:T659,### Sources “Six-Year Review 4 Compliance Monitoring Data (2012-2019) | US EPA.” SYR4_PhaseChem_3.zip. US EPA, 29 Feb. 2024, https://www.epa.gov/dwsixyearreview/six-year-review-4-compliance-monitoring-data-2012-2019. “Data Downloads | ECHO | US EPA.” SDWA Dataset (ZIP), SDWA_GEOGRAPHIC_AREAS.csv. Echo.Epa.Gov, https://echo.epa.gov/tools/data-downloads. Accessed 25 Sept. 2026. US Census Bureau. “Gazetteer Files.” National Counties Gazetteer File. Census.Gov, 4 Sept. 2026, https://www.census.gov/geographies/reference-files/time-series/geo/gazetteer-files.html. "Prohibition of fluoride in Utah's public water systems." HB-81. 2025, https://le.utah.gov/Session/2025/bills/enrolled/HB0081.pdf. Taylor, Kyla W, et al. “Fluoride Exposure and Children’s IQ Scores.” JAMA Pediatrics, vol. 179, no. 3, American Medical Association, Jan. 2025, https://doi.org/10.1001/jamapediatrics.2024.5542. “Fluoride Exposure: Neurodevelopment and Cognition.” National Toxicology Program, 23 Apr. 2025, https://ntp.niehs.nih.gov/research/assessments/noncancer/completed/fluoride. Aoun, Antoine, et al. “The Fluoride Debate: The Pros and Cons of Fluoridation.” Preventive Nutrition and Food Science, vol. 23, no. 3, Sept. 2018, pp. 171–80, https://doi.org/10.3746/pnf.2018.23.3.171. Cleveland Clinic. “Fluorosis: What It Is, Causes & Treatment.” Cleveland Clinic, 8 June 2022, https://my.clevelandclinic.org/health/diseases/23227-fluorosis. ### Ai Citation https://share.gemini.google/01mQqEOVo1dJ https://share.gemini.google/6RlTlJgDzsko https://claude.ai/share/bba37212-7a73-4e32-97b0-883da8798d74 33:[{"id":"40a9e8b08f2a2740a4b10d84f6914b07603b8fa49d7c45a8d0c8ccbafae1ac32fb825916202f679b448af7e8d5e67714d5c2896740288d2d5d387dfd679a0b64","href":"https://static.observableusercontent.com/files/40a9e8b08f2a2740a4b10d84f6914b07603b8fa49d7c45a8d0c8ccbafae1ac32fb825916202f679b448af7e8d5e67714d5c2896740288d2d5d387dfd679a0b64","name":"Method.png","type":"image/png","lastModified":1790462966318,"size":54269},{"id":"3059572a683abbb8cc6c5c0dcfd9d543c83c4ca50d55b3697dea4190f6d69453d1df302bf9be8a1b1c0d1c13f95234da1a490d8023d40b7d7487d43e91eb5929","href":"https://static.observableusercontent.com/files/3059572a683abbb8cc6c5c0dcfd9d543c83c4ca50d55b3697dea4190f6d69453d1df302bf9be8a1b1c0d1c13f95234da1a490d8023d40b7d7487d43e91eb5929","name":"PERFECT_WATER_FLUORIDE.csv","type":"text/csv","lastModified":1790463954064,"size":34445142},{"id":"8ae75e34d9bfc07a926d418181e2bc0b870b4382a563cd84c4fdf5d4e12387ebd199898c6e3ad8f6e5596710ece88c6c61ea71a654023a8f92856f7c63aa44ab","href":"https://static.observableusercontent.com/files/8ae75e34d9bfc07a926d418181e2bc0b870b4382a563cd84c4fdf5d4e12387ebd199898c6e3ad8f6e5596710ece88c6c61ea71a654023a8f92856f7c63aa44ab","name":"IMG_0704.webp","type":"image/webp","lastModified":1790592829272,"size":55524}] 2c:{"id":"5047298b42ec4ee8","slug":"fluoride-analysis","title":"Should we fear fluoride?","thumbnail":"e9df9691e301eb17893eaf5abc30e4948bc05eb013787b5d570646e99dfe4a3e","access_level":"public","create_time":"$D2026-09-18T17:16:03.102Z","update_time":"$D2026-09-28T16:58:24.780Z","view_time":null,"first_public_time":"$D2026-09-28T12:19:27.422Z","has_unpublished_edits":null,"fork":false,"stars":1,"starred":false,"owner":{"type":"team","id":"fbf9eef530f9c2a1","login":"grant-m","name":"Grant M","avatar_url":"https://avatars-next.observableusercontent.com/2ffa116de23fbd0c0705acf3f6c2979c61cbad756e62c014e71a48ebad16d37a","tier":"starter_2024"},"version":779,"publish_version":779,"role":null,"authored":false,"authors":[{"type":"individual","id":"7dbd1e6e69c36ea1","login":"grant-m","name":"Grant M","avatar_url":"https://avatars-next.observableusercontent.com/2ffa116de23fbd0c0705acf3f6c2979c61cbad756e62c014e71a48ebad16d37a","tier":null}],"members":[{"type":"team","id":"fbf9eef530f9c2a1","login":"grant-m","name":"Grant M","avatar_url":"https://avatars-next.observableusercontent.com/2ffa116de23fbd0c0705acf3f6c2979c61cbad756e62c014e71a48ebad16d37a","tier":"starter_2024","role":"editor"}],"comments":[],"custom_thumbnail":false,"trashed":false,"lapsed":null,"subscribed":null,"forks":0,"collection":null,"type":"notebook","body":{"title":"Should we fear fluoride?","cells":[{"id":1,"value":"# Should we fear fluoride?","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":72,"value":"

\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":8,"value":"$2e","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":2,"value":"### Hypothesis\n\nThe water supplies of most reasonably large American communities do not contain a harmful amount of fluoride.\n","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":9,"value":"### Method","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":50,"value":"We refer to the Six-Year Review 4 Compliance Monitoring Data (2012-2019) by the EPA, which regularly samples how much fluoride is in the water supply at given times. Large amounts of data manipulation are necessary, as this database does not directly contain geographical information. Instead, we reference the PSWID of each water facility along with their County code in order to create a Geographic ID. This can be then be queried within the US Census Bureau Gazetteer files in order to access an approximate latitude and longitude score for each facility. It should be noted that some databases include geographic information to their legal representative, which is not useful for our purposes and presumably the purposes of the reader. It should be further noted that this process is not universal and not all PSWIDs were able to be linked to a Geographic ID via our method. Our solution to this was to drop such data, which is regrettable chunk, but likely does not have significant consequences for our method.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":12,"value":"FileAttachment(\"Method.png\").image()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":52,"value":"Here is a sample of the database we are working on. It can be separated into two parts, identifying data and valuable data. We have removed PSWID and Geographic IDs from the database as it is not relevant to our analysis and the water facility ID should be more than enough for extraordinary research ideas.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":14,"value":"database = FileAttachment(\"PERFECT_WATER_FLUORIDE.csv\").csv()","pinned":false,"hidden":false,"mode":"ojs","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":49,"value":"Inputs.table([0,5287,9951,59421,211204,311831].map(x=>database[x]), {select: false})","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":54,"value":"Currently, the U.S. Public Health Service sets 0.7 mg/L as a safe upper bound for people of all ages. However, there is a lack of evidence if this is a tight a bound for adults, as this bound was likely set with respect to Taylor, 2025. The World Health Organization provides an alternative bound, citing 1.5 mg/L as a safe level of fluoride (National Toxicology Program 2025). Thus, we will be looking for large quantities of communities which oppose this rule.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":10,"value":"### Results","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":53,"value":"We visualize the database in the following figure. Water samples are grouped by year as otherwise, overlapping data becomes far too apparent in order to reasonably parse. Areas that serve over a million people are arguably more relevant to this study, so their scale has been sized up with respect to their population.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":45,"value":"const ydatabase = Object.values(\n Object.groupBy(\n database.filter(item => item.SAMPLE_COLLECTION_DATE?.split(\"-\")?.[2]),\n (item) => item.SAMPLE_COLLECTION_DATE.split(\"-\")[2]\n )\n);\n\nconst yearmap = {\"2019\":7, \"2018\":6, \"2017\":5, \"2016\":4, \"2015\":3, \"2014\":2, \"2013\":1, \"2012\":0};\n\nconst select_year = view(\n Inputs.radio([\"2019\", \"2018\", \"2017\", \"2016\", \"2015\", \"2014\", \"2013\", \"2012\"], {label: \"Select Year\", value: \"2019\"})\n);","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":65,"value":"const title = Plot.plot({title: `${select_year} Fluoride levels in water by US community`});\ndisplay(title);","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":47,"value":"$2f","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":67,"value":"The figure uses some tight clustering and some clever logic with the tip plots in order to produce what I think is a pretty understandable figure. As stated above, it breaks down the years into separate cases per year. While this does lead to some pretty tightly clustered data, especially on the east coast, shrinking it down to smaller intervals makes this figure less user friendly. Not to mention, you risk losing more data to make the figure pretty interpretable.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":11,"value":"### Analysis","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":68,"value":"For further study, this is a supplementary tool that allows a user to control the exact amounts of fluoride and population limits. The purpose of this tool is because we realize different users will have different values. For example, some users may believe that smaller communities may have not yet adopted a modern fluoridation or water filtering system. Or, some users may believe that a 0.7 limit is far too strict, and only care about the more liberal 1.5 limit. Thus, the sliders below will assist in complementary analysis as well as providing a small statistical breakdown of the filters affect on the data. Please do note that this figure uses means in calculations, which may introduce error.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":62,"value":"const means_min = Array.from(\n d3.rollup(\n database,\n v => ({\n name: `${v[0].SYSTEM_NAME} ${v[0].STATE_CODE}`,\n lat: +v[0].INTPTLAT,\n lon: +v[0].INTPTLONG,\n meanFluoride: d3.mean(v, d => +d[\"VALUE MG/L\"] > 0.7),\n maxFluoride: d3.max(v, d => +d[\"VALUE MG/L\"]),\n meanPopulation: d3.mean(v, d => +d[\"ADJUSTED_TOTAL_POPULATION_SERVED\"]),\n n: v.length\n }),\n d => `${d.STATE_CODE}|${d.INTPTLAT}|${d.INTPTLONG}`\n ).values()\n);\n\nconst countf = means_min.filter(d =>\n +d.meanFluoride > +fluoride_limit && +d.meanPopulation > +population_limit\n).length;\n\nconst ppl_sum = means_min.filter(d =>\n +d.meanFluoride > +fluoride_limit && +d.meanPopulation > +population_limit\n).reduce((sum, row) => sum + row.meanPopulation, 0);","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":63,"value":"const fluoride_limit = view(Inputs.range([0, 1.5], {\n step: 0.01,\n value: 0.7,\n label: \"Fluoride Limit\",\n}));\n\nconst population_limit = view(Inputs.range([0, 400000], {\n step: 100,\n value: 1000,\n label: \"Population Limit\",\n}));\n","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":60,"value":"$30","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":70,"value":"$31","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":4,"value":"$32","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":15,"value":"### Procedure","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":57,"value":"FileAttachment(\"IMG_0704.webp\").image()","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":58,"value":"This is a preliminary sketch of the rough idea of how I want to introduce this. I am not sure that a min-max range is a good way to go, because I don't really care about how much fluoride we can get in the water, but rather what areas are exceeding certain limits. This needs some toggability or something, I haven't really planned out user interactions at this stage.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":18,"value":"import us from \"npm:us-atlas/states-10m.json\" with {type: \"json\"};","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":32,"value":"const locations = Array.from(\n d3.rollup(\n database,\n v => ({\n lat: +v[0].INTPTLAT,\n lon: +v[0].INTPTLONG,\n state: v[0].STATE_CODE,\n meanFluoride: d3.mean(v, d => +d[\"VALUE MG/L\"]),\n maxFluoride: d3.max(v, d => +d[\"VALUE MG/L\"]),\n n: v.length\n }),\n d => `${d.STATE_CODE}|${d.INTPTLAT}|${d.INTPTLONG}`\n ).values()\n);","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":34,"value":"const plot = Plot.plot({\n width: 975,\n height: 610,\n projection: \"albers-usa\",\n color: {\n type: \"linear\",\n scheme: \"turbo\",\n domain: [0, 2],\n label: \"Mean fluoride (mg/L)\",\n legend: true,\n clamp: true\n },\n title: \"Fluoride Levels in U.S. Public Water Systems\",\n subtitle: \"Mean fluoride concentration (mg/L) by water system location, 2012–2019 compliance monitoring data\",\n marks: [\n Plot.geo(statesGeo, {fill: \"#eee\", stroke: \"#fff\"}),\n Plot.dot(locations.filter(d => d.state !== \"\"), {\n x:\"lon\",\n y:\"lat\",\n r: \"meanFluoride\",\n fill: \"meanFluoride\",\n tip:true}),\n\n ]\n});\ndisplay(plot);\n","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":21,"value":"I went ahead and used observables built-in Ai to generate this by just asking it to generate a plot off of my database. I'm actually pretty impressed with how it turned out. I am not sure about a mean approach to this, though it is certainly not a bad idea. I think the main issue is that it refuses to take population size into account. Most of these data points are relatively small, so Texas looks awful. Also, it's not interactable for the user.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":22,"value":"const statesGeo = topojson.feature(us, us.objects.states);\n\nconst plot = Plot.plot({\n width: 975,\n height: 610,\n projection: \"albers-usa\",\n color: {\n type: \"linear\",\n scheme: \"turbo\",\n domain: [0, 2],\n label: \"Mean fluoride (mg/L)\",\n legend: true,\n clamp: true\n },\n title: \"Fluoride Levels in U.S. Public Water Systems\",\n subtitle: \"Mean fluoride concentration (mg/L) by water system location, 2012–2019 compliance monitoring data\",\n marks: [\n Plot.geo(statesGeo, {fill: \"#eee\", stroke: \"#fff\"}),\n Plot.density(locations.filter(d => d.state !== \"\"), {\n x: \"lon\",\n y: \"lat\",\n bandwidth: 10,\n fill: \"density\",\n r: \"meanFluoride\",\n title: d => `${d.state}\\nmean: ${d.meanFluoride.toFixed(2)} mg/L\\nmax: ${d.maxFluoride.toFixed(2)} mg/L\\nsamples: ${d.n}`,\n tip: true\n }),\n Plot.dot(locations.filter(d => d.state !== \"\"), {x:\"lon\", y:\"lat\", r: 1, fill: \"currentColor\"}),\n\n ]\n});\ndisplay(plot);\n","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":24,"value":"I've been messing with getting a heatmap display working, but I think this counts as a raster. Since we aren't supposed to do those, I'll pivot to just using points. Also, this just shows the density of sampling and doesn't have any impact on the actual fluoride data itself.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":25,"value":"function parseSampleDate(s) {\n if (!s) return null;\n const d = new Date(s.replace(/-/g, \" \"));\n return Number.isNaN(+d) ? null : d;\n}\n\nconst parsedDatabase = database.map(d => ({\n ...d,\n date: parseSampleDate(d.SAMPLE_COLLECTION_DATE)\n})).filter(d => d.date);\n\nconst uniqueDates = Array.from(\n new Set(parsedDatabase.map(d => +d.date))\n).sort((a, b) => a - b).map(t => new Date(t));\n\nuniqueDates.length;","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":26,"value":"const monthBins = Array.from(\n new Set(parsedDatabase.map(d => +d3.timeMonth.floor(d.date)))\n).sort((a, b) => a - b).map(t => new Date(t));\n\nmonthBins.length;","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":30,"value":"We separate our data out into bins in order to create a slider so the user can investigate how the data changes over time.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":28,"value":"const selectedMonth = monthBins[monthIndex];\n\nconst filteredDatabase = parsedDatabase.filter(\n d => +d3.timeMonth.floor(d.date) === +selectedMonth\n);\n\nconst locationsByMonth = Array.from(\n d3.rollup(\n filteredDatabase,\n v => ({\n lat: +v[0].INTPTLAT,\n lon: +v[0].INTPTLONG,\n state: v[0].STATE_CODE,\n meanFluoride: d3.mean(v, d => +d[\"VALUE MG/L\"]),\n maxFluoride: d3.max(v, d => +d[\"VALUE MG/L\"]),\n n: v.length\n }),\n d => `${d.STATE_CODE}|${d.INTPTLAT}|${d.INTPTLONG}`\n ).values()\n).filter(d => Number.isFinite(d.lat) && Number.isFinite(d.lon) && d.state !== \"\");\n\nlocationsByMonth.length;","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":37,"value":"const monthIndex = view(Inputs.range([0, monthBins.length - 1], {\n step: 1,\n value: 0,\n label: \"Sample month\",\n format: i => d3.timeFormat(\"%B %Y\")(monthBins[i])\n}));","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":29,"value":"const statesGeoByMonth = topojson.feature(us, us.objects.states);\n\nconst monthPlot = Plot.plot({\n width: 975,\n height: 610,\n projection: \"albers-usa\",\n color: {\n type: \"linear\",\n scheme: \"turbo\",\n domain: [0, 2],\n label: \"Mean fluoride (mg/L)\",\n legend: true,\n clamp: true\n },\n title: \"Fluoride Levels in U.S. Public Water Systems\",\n subtitle: `Mean fluoride concentration (mg/L) by water system location — ${d3.timeFormat(\"%B %Y\")(selectedMonth)}`,\n marks: [\n Plot.geo(statesGeoByMonth, {fill: \"#eee\", stroke: \"#fff\"}),\n Plot.dot(locationsByMonth, {\n x: \"lon\",\n y: \"lat\",\n r: \"n\",\n fill: \"meanFluoride\",\n fillOpacity: 0.75,\n stroke: \"black\",\n strokeOpacity: 0.2,\n title: d => `${d.state}\\nmean: ${d.meanFluoride.toFixed(2)} mg/L\\nmax: ${d.maxFluoride.toFixed(2)} mg/L\\nsamples: ${d.n}`,\n tip: true\n })\n ]\n});\ndisplay(monthPlot);","pinned":false,"hidden":false,"mode":"js","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":35,"value":"I want to encode time as well, but month is way to laggy. It's got a cool effect on the circles, but I'm not sure if this encompassing all the data well enough to really support my hypothesis well. Also it'd be good to build my fluoride limits into the figure.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":39,"value":"const yearDatabase = Object.values(\n Object.groupBy(\n database.filter(item => item.SAMPLE_COLLECTION_DATE?.split(\"-\")?.[2]),\n (item) => item.SAMPLE_COLLECTION_DATE.split(\"-\")[2]\n )\n);\n\ndisplay(yearDatabase);","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":36,"value":"const yearIndex = view(\n Inputs.range([0, yearDatabase.length - 1], {\n value: 0,\n step: 1,\n title: \"Sample Year\"\n })\n);\n\ndisplay(yearIndex)","pinned":false,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":42,"value":"const yearPlot = Plot.plot({\n width: 975,\n height: 610,\n projection: \"albers-usa\",\n color: {\n type: \"linear\",\n scheme: \"turbo\",\n domain: [0, 2],\n label: \"VALUE MG/L (mg/L)\",\n clamp: true\n },\n title: \"Fluoride Levels in U.S. Public Water Systems\",\nmarks: [\n Plot.geo(statesGeoByMonth, {fill: \"#eee\", stroke: \"#fff\"}),\n Plot.dot(yearDatabase[yearIndex], {\n x: \"INTPTLONG\",\n y: \"INTPTLAT\"\n })\n]\n});\ndisplay(yearPlot);","pinned":true,"hidden":false,"mode":"ts","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"},{"id":44,"value":"With a working year slider, all that's left to do is put it all together and add some colors to make it look pretty. Since we have some large clusters, I might also only take the largest amount of fluoride that's sampled per year. I could also look at half years rather than full years to (hopefully) render less data.","pinned":false,"hidden":false,"mode":"md","author":"$undefined","summary":"$undefined","output":"$undefined","database":"$undefined"}],"files":"$W33","stdlib":"2","titling":"auto"}} 2d:"$2c:body"