Published
Edited
Oct 5, 2022
1 fork
Importers
1 star
ss.sampleCorrelation(
data.map((item) => item.x1),
data.map((item) => item.y1)
)
dataLong = {
const arrayTable = data.map((item) => Object.values(item)); //convert to an array (removing column headings)
const resultTable = [];
for (let i = 0; i < 4; i++) {
for (let rowArray of arrayTable) {
resultTable.push({
x: rowArray[i * 2],
y: rowArray[i * 2 + 1],
set: i + 1
});
}
}
return resultTable;
}
T = require("@tidyjs/tidy/dist/umd/tidy.min.js")
ss = require("simple-statistics")
summaryStatistics = T.tidy(
dataLong,
T.groupBy(
"set",
T.summarize({
//TidyJs's built-in T.mean() function, in this case that finds the mean of all y values in the subset
yMean: T.mean("y"),
xMean: T.mean("x"),
yMean: T.mean("y"),
xVariance: T.variance("x"),
yVariance: T.variance("y"),
//Here, I've written an inline lamba function (a one liner) where TidyJS provides the subset of rows
//and I extract the xs and ys as required for simple-statistics' sampleCorrelation
//function https://simplestatistics.org/docs/#sampleCorrelation
correlation: (subset) =>
ss.sampleCorrelation(
subset.map((item) => item.x),
subset.map((item) => item.y)
),
//linear regression grident
regressionM: (subset) =>
ss.linearRegression(subset.map((item) => [item.x, item.y])).m,
//linear regression y intercept
regressionB: (subset) =>
ss.linearRegression(subset.map((item) => [item.x, item.y])).b,
//Here, I've written a multi-line inline function (between braces with a return statement)
//which calculates the rSquared value function https://simplestatistics.org/docs/#rSquared
rSquared: (subset) => {
const samples = subset.map((item) => [item.x, item.y]);
const regressionLine = ss.linearRegressionLine(
ss.linearRegression(samples)
);
return ss.rSquared(samples, regressionLine);
}
})
)
)
Inputs.table(summaryStatistics)
vlEmbed = require("vega-embed")
summaryStatisticsLonger = T.tidy(
summaryStatistics,
T.pivotLonger({ cols: "-set", namesTo: "statistic", valuesTo: "value" })
)
summaryStatsChart=vlEmbed({
data: { values: summaryStatisticsLonger },
mark: "bar",
encoding: {
y: { field: "value", type: "quantitative" },
x: { field: "set" },
facet: { field: "statistic" },
color: { field: "set" }
}
})
vlEmbed({
data: { values: dataLong },
mark: "point",
encoding: {
x: { field: "x", type: "quantitative" },
y: { field: "y", type: "quantitative" },
facet: { field: "set" },
color: { field: "set" }
}
})
moreSummaryStatisticsLonger = T.tidy(
moreSummaryStatistics,
T.pivotLonger({ cols: "-set", namesTo: "statistic", valuesTo: "value" })
)
vlEmbed({
data: { values: moreSummaryStatisticsLonger },
mark: "bar",
encoding: {
y: { field: "value", type: "quantitative" },
x: { field: "set" },
facet: { field: "statistic" },
color: { field: "set" }
}
})
vl
.markBar()
.data(summaryStatisticsLonger)
.encode(
vl.x().fieldO("set"),
vl.y().fieldQ("value"),
vl.facet().fieldO("statistic"),
vl.color().fieldN("set")
)
.render()
vl
.markPoint()
.data(dataLong)
.encode(
vl.x().fieldQ("x"),
vl.y().fieldQ("y"),
vl.facet().fieldO("set"),
vl.color().fieldN("set")
)
.render()
Purpose-built for displays of data
Observable is your go-to platform for exploring data and creating expressive data visualizations. Use reactive JavaScript notebooks for prototyping and a collaborative canvas for visual data exploration and dashboard creation.
