locutus/esm/python/statistics/median_grouped.js

Locutus other languages' standard libraries to JavaScript for fun and educational purposes

import { assertStatisticsArray, sortStatisticsValues, toFloatStatisticNumber } from "../_helpers/_statistics.js";
export function median_grouped(data, interval = 1) {
    //      discuss at: https://locutus.io/python/statistics/median_grouped/
    // parity verified: Python 3.12
    //     original by: Kevin van Zonneveld (https://kvz.io)
    //          note 1: Estimates the median for data binned around interval midpoints using grouped interpolation.
    //       example 1: median_grouped([52, 52, 53, 54])
    //       returns 1: 52.5
    //       example 2: median_grouped([52, 52, 53, 54], 2)
    //       returns 2: 52
    //       example 3: median_grouped([true, false, true])
    //       returns 3: 0.75
    const values = sortStatisticsValues(assertStatisticsArray(data, 'median_grouped'), 'median_grouped');
    const n = values.length;
    if (n === 0) {
        throw new Error('no median for empty data');
    }
    const x = values[Math.floor(n / 2)];
    let i = 0;
    while (i < values.length && values[i] !== x) {
        i += 1;
    }
    let j = i;
    while (j < values.length && values[j] === x) {
        j += 1;
    }
    const numericInterval = toFloatStatisticNumber(interval, 'median_grouped');
    const numericX = toFloatStatisticNumber(x, 'median_grouped');
    const lowerLimit = numericX - numericInterval / 2;
    const cumulativeFrequency = i;
    const frequency = j - i;
    return lowerLimit + (numericInterval * (n / 2 - cumulativeFrequency)) / frequency;
}