estimate-tasks

About

Given the individual tasks' estimations (in hours), calculates the total estimation for the whole project. It attempts to do the right thing, when summing up estimations, treating each individual estimate as a continuous random variable (see Math), and applying a formula from Steve McConnel's book "Software Estimation: Demystifying the Black Art".

Doing the math right though requires providing a range for an individual estimate instead of a single number: you'll be asked for the best case, the most likely case, and the worst case hours.

By the way, the probability of a task taking exactly X hours to finish is exactly zero (see Math again ;).

Also, a subjective confidence factor is provided for each task's estimation, in percent, within the range (0, 100) -- excluding 0 and 100. Normally, you should be pretty confident in the most of your tasks' estimations, specifying confidence factors above 90% for them. Only a smaller part of tasks should have 90% and below. Anyway, that's a just fine tuning mechanism; you may well ignore it, and simply assign, say, 95% confidence to all the tasks.

The number E you get from total75PercentLikelyHours() means that there's a 75% probability that your project will take not more than E hours.

Installation

Node.js:

$ npm install estimate-tasks

Web browser:

$ bower install estimate-tasks

API

Node version:

var et = require("estimate-tasks"),

    total = et.total75PercentLikelyHours([{
        bestCaseHours: 1,
        mostLikelyCaseHours: 1.5,
        worstCaseHours: 4,
        confidencePercent: 99.9
    }, {
        bestCaseHours: 8,
        mostLikelyCaseHours: 10,
        worstCaseHours: 12,
        confidencePercent: 66
    }, {
        bestCaseHours: 2,
        mostLikelyCaseHours: 3,
        worstCaseHours: 4,
        confidencePercent: 95
    }]);

In the browser you have to include the lib/calc.js script. Then:

var et = window["estimate-tasks"];

Testing

Test-driven with Mocha. Under Node, say:

~/estimate-tasks(master)$ npm test