Bayesian Prior Calculator

This calculator applies Bayes' theorem (line 8: P(A|B) = P(B|A)P(A) / [P(B|A)P(A) + P(B|¬A)P(¬A)]) to update a Prior Probability P(A) after a positive or negative test result, using Sensitivity as the true-positive rate and False Positive Rate as the false-alarm rate. All three inputs pass through Math.min(Math.max(inputs.X || default, 0.0001), 0.9999) (lines 4-6) before any arithmetic runs — but the || runs first, so 0 and 1 aren't treated the same way. Typing exactly 1 is truthy, so it reaches the clamp untouched and gets pulled down to 0.9999. Typing exactly 0 (or clearing the field) is falsy, so it's replaced by that field's own default — 0.01 for Prior Probability, 0.9 for Sensitivity, 0.05 for False Positive Rate — before the clamp ever runs, so it isn't clamped to a boundary at all.

Positive Likelihood Ratio and Negative Likelihood Ratio (lines 25-26) are pure ratios of Sensitivity and False Positive Rate — Prior Probability plays no role in either formula at all, a fact readable directly from the code rather than inferred from a sweep. Prior Odds (line 29) is likewise computed purely from Prior Probability and stays completely fixed however Sensitivity or False Positive Rate change. Posterior P(A | negative test) responds far more strongly to Sensitivity than to Prior Probability, since a highly sensitive test is defined by how rarely it produces a false negative — exactly what a negative result needs to rule out. This calculator does not account for sequential testing: running the same test twice, or combining two different tests, requires manually feeding the first posterior back in as the new prior, since there's no built-in mechanism here for chaining test results.