Published
Edited
Apr 4, 2022
1 fork
14 stars
Plot: Stacked bin dotsWaffle ChartRenko ChartHow to attach a plot to a DOM elementIrregular Bar ChartDatabase Outage PlotLog-scale histogramIrregular bins histogramPlot HyperboloidHappy anniversary, Project Gutenberg!ELD ViewerMatrix diagramK-means binningHello, Open MeteoRoad traffic under COVID restrictions (France)How many books was I reading?Metbrewer color schemesIrregular scalesTopological subsamplingVariants du SARS-CoV-2 en FranceVariants of SARS-Cov-2 in EuropeVariable tick colorMinimalist dodge plotAssemblée nationale 2017-2022AoC 12: shortest path under constraintsHow fast does walk-on-spheres converge?
ROC curve
result = `${testname}_result`
roc = d3.sort(d3.shuffle(data.filter((d) => d[result] >= 0)), (d) => d[result])
threshold = (roc[threshold_i][result] + roc[threshold_i + 1][result]) / 2
// this is the threshold value where the test's interpretation changes (from "NOT detected" to "DETECTED")
threshold_i = {
const interpretation = `${testname}_interpretation`;
for (let i = 0; i < roc.length - 1; i++)
if (roc[i][interpretation] !== roc[i + 1][interpretation]) return i;
// for random and ideal, set the threshold at the number of true negative
return N - 1;
}
T = Uint8Array.from(roc, (d) => +(d.status === "P"))
X = d3.cumsum(T)
P = X[T.length - 1] // equiv. d3.sum(T)
Y = X.map((x, i) => 1 + i - x) // equiv. d3.cumsum(T, t => 1 - t)
N = Y[T.length - 1] // equiv. d3.sum(T, t => 1 - t)
AUC = 1 - d3.sum(T, (t, i) => (1 - t) * X[i]) / (N * P)
data = FileAttachment("supplementary_dataset2.xlsx")
.xlsx()
.then((xl) => xl.sheet("all_samples", { headers: true }))
.then(
(data) => (
data.forEach((d) => {
d.status = d.samplecollection === "obb_negatives" ? "N" : "P";
d.status_result = 1 + +(d.status === "P");
d.diasorin_result =
d.diasorin_result[0] === "<"
? +d.diasorin_result.slice(1).trim()
: d.diasorin_result;
d.siemens_result =
d.siemens_result[0] === "<"
? +d.siemens_result.slice(1).trim()
: d.siemens_result;
d.random_result = random();
}),
data
)
)
random = d3.randomExponential(1)
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.