Public
Edited
Sep 2, 2024
1 fork
// Going from match weight to probability
mw_1 = {
const rounded_partial_match_weight = roundToSignificantFigures(
selected_val,
4
);
return html`${tex`\text{Partial Match Weight} = \omega = ${rounded_partial_match_weight}`}`;
}
bf_1 = {
const bayes_factor = roundToSignificantFigures(
calcBayesFactorFromPartialMatchWeight(selected_val),
4
);
const p_val = calcProbabilityFromPartialMatchWeight(selected_val).toFixed(4);
const rounded_partial_match_weight = roundToSignificantFigures(
selected_val,
4
);
return html`${tex`\text{Bayes Factor} = K = 2^{\omega} = 2^{{${rounded_partial_match_weight}}} = ${bayes_factor}`}`;
}
p_1 = {
const p_val = calcProbabilityFromPartialMatchWeight(selected_val).toFixed(4);
const partial_match_weight = roundToSignificantFigures(selected_val, 4);
return html`${tex`\text{Probability} = p = \frac{K}{1 + K} = \frac{2^{{${partial_match_weight}}}}{1 + 2^{{${partial_match_weight}}}} = ${p_val}`}`;
}
html`${tex`\text{Bayes Factor} = K = \frac{p}{1-p}`}`
// Going from probability to match weight
p_2 = {
const p_val = calcProbabilityFromPartialMatchWeight(selected_val);
const p_val_rounded = p_val.toFixed(4);
return html`${tex`\text{Probability} = p = ${p_val_rounded}`}`;
}
bf_2 = {
const p_val = calcProbabilityFromPartialMatchWeight(selected_val);
const p_val_rounded = p_val.toFixed(4);
const bayes_factor = roundToSignificantFigures(
calcBayesFactorFromProbability(p_val),
4
);
return html`${tex`\text{Bayes Factor} = K = \frac{p}{1-p} = \frac{${p_val_rounded}}{1 - ${p_val_rounded}} = ${bayes_factor}`}`;
}
viewof prior_probabability = Inputs.range([0, 1], {
label: "Prior probability"
})
viewof max_match_weight = Inputs.range([5, 40], {
label: "Max match weight",
step: 5,
value: 10 // Default value
})
// https://observablehq.com/@visnup/vega-lite-data-out?collection=@visnup/visualization
// https://observablehq.com/@robinl/m-and-u-probabilities
function probToBayesFactor(prob) {
return prob !== 1 ? prob / (1 - prob) : Infinity;
}
function probToMatchWeight(prob) {
const bayesFactor = probToBayesFactor(prob);
const matchWeight = Math.log2(bayesFactor);
// Round to avoid floating point errors
const precision = 8; // You can adjust the precision as needed
return (
Math.round(matchWeight * Math.pow(10, precision)) / Math.pow(10, precision)
);
}
function matchWeightToBayesFactor(weight) {
return Math.pow(2, weight);
}
function bayesFactorToProb(bf) {
return bf / (1 + bf);
}
function formatBf(bf, mw) {
if (bf >= 1000) {
return bf.toLocaleString(undefined, { maximumFractionDigits: 0 });
} else if (bf >= 100) {
return bf.toLocaleString(undefined, { maximumFractionDigits: 1 });
} else if (bf >= 10) {
return bf.toLocaleString(undefined, { maximumFractionDigits: 2 });
} else if (bf >= 1) {
return bf
.toLocaleString(undefined, { maximumFractionDigits: 3 })
.replace(/\.?0+$/, "");
} else {
// Apply similar formatting logic as prob_fmt for bf < 1
const absMw = Math.abs(mw);
const additionalDigits = Math.floor(absMw / 5);
const minFractionDigits = 4 + additionalDigits;
return bf
.toLocaleString(undefined, { minimumFractionDigits: minFractionDigits })
.replace(/\.?0+$/, "");
}
}
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.