Related
You need to find some unknown, predetermined point in three-dimensional space, in the smallest number of attempts, using only a function that can return the distance from
any point you pass to it to the desired unknown point.
To solve the problem, first implement a function f that, by taking the coordinates of any point s(x, y, z), return the distance between that point and a conditionally unknown, randomly generated point
point you arbitrarily generate r(x, y, z), where x, y, z can be integers between
0 и 100.
For example, for an arbitrarily generated point r(0, 0, 10) and a point passed to the function
s(0, 0, 0), the result of the function would be as follows:
f(s) = 10 // the distance between s(0, 0, 0) and r(0, 0, 10) is 10
Next, implement the algorithm itself for the assignment. The algorithm should find the coordinates of
of an arbitrarily generated point with the least number of calls to the function f.
I have a randomizer instead of an algorithm, that's all I got. Help.
const pointToFound = {
x: 12,
y: 9,
z: 76,
};
let attemts = 0;
let isXFound = false;
let isYFound = false;
let isZFound = false;
const pointHistory = [];
const getRandomPoint = () => {
return {
x: isXFound ? isXFound : Math.floor(Math.random() * 101),
y: isYFound ? isYFound : Math.floor(Math.random() * 101),
z: isZFound ? isZFound : Math.floor(Math.random() * 101),
};
};
const getDifference = (point, pointToCompare) => {
return {
x:
Math.max(point.x, pointToCompare.x) - Math.min(point.x, pointToCompare.x),
y:
Math.max(point.y, pointToCompare.y) - Math.min(point.y, pointToCompare.y),
z:
Math.max(point.z, pointToCompare.z) - Math.min(point.z, pointToCompare.z),
};
};
const condition = !isXFound && !isYFound && !isZFound;
while (condition) {
const point = getRandomPoint();
const difference = getDifference(point, pointToFound);
pointHistory.push(point);
attemts += 1;
if (isXFound && isYFound && isZFound) {
console.log("Total attempts: ", attemts);
console.log(point);
break;
}
if (difference.x === 0 && !isXFound) {
isXFound = point.x;
}
if (difference.y === 0 && !isYFound) {
isYFound = point.y;
}
if (difference.z === 0 && !isZFound) {
isZFound = point.z;
}
}
console.log(pointHistory);
I have a randomizer instead of an algorithm, that's all I got. Help.
This can be done with at most 3 guesses and often with 2 guesses:
Let the first guess be [0, 0, 0], and ask for the distance
Find in the 100x100x100 cube all points that have that distance to [0, 0, 0]. There might be around 100-200 points that have that distance: consider all of these candidates.
Take the first candidate as the second guess and ask for the distance
Find among the other candidates the ones that have exactly that distance to the first candidate. Often there will be only one point that satisfies this condition. In that case we can return that candidate and only 2 guesses were necessary.
Otherwise (when there is more than one candidate remaining) repeat the previous step which will now certainly lead to a single point.
Here is an implementation that provides a blackbox function which chooses the secret point in a local variable, and which returns two functions: f for the caller to submit a guess, and report for the caller to verify the result of the algorithm and report on the number of guesses. This is not part of the algorithm itself, which is provided in the findPoint function.
const rnd = () => Math.floor(Math.random() * 101);
const distance = (a, b) =>
a.reduce((acc, x, i) => acc + (x - b[i]) ** 2, 0) ** 0.5;
function findPoint(f) {
// First guess is always the zero-point
let guess = [0, 0, 0];
let dist = f(guess);
if (dist === 0) return guess; // Extremely lucky!
// Find the points in the cube that have this distance to [0,0,0]
let candidates = [];
const limit = Math.min(100, Math.round(dist));
for (let x = 0; x <= limit; x++) {
const p = [x, limit, 0];
// Follow circle in X=x plane
while (p[1] >= 0 && p[2] <= limit) {
const d = distance(p, guess);
const diff = d - dist;
if (Math.abs(diff) < 1e-7) candidates.push([...p]);
if (diff >= 0) p[1]--;
else p[2]++;
}
}
// As long there are multiple candidates, continue with a guess
while (candidates.length > 1) {
const candidates2 = [];
// These guesses are taking the first candidate as guess
guess = candidates[0];
dist = f(guess);
if (dist === 0) return guess; // lucky!
for (const p of candidates) {
let d = distance(p, guess);
let diff = d - dist;
if (Math.abs(diff) < 1e-7) candidates2.push(p);
}
candidates = candidates2;
}
return candidates[0]; // Don't call f as we are sure!
}
function blackbox() {
const secret = [rnd(), rnd(), rnd()];
console.log("Secret", JSON.stringify(secret));
let guessCount = 0;
const f = guess => {
guessCount++;
const dist = distance(secret, guess);
console.log("Submitted guess " + JSON.stringify(guess) + " is at distance " + dist);
return dist;
};
const report = (result) => {
console.log("Number of guesses: " + guessCount);
console.log("The provided result is " + (distance(secret, result) ? "not" : "") + "correct");
}
return {f, report};
}
// Example run
const {f, report} = blackbox();
const result = findPoint(f);
console.log("Algorithm says the secret point is: " + JSON.stringify(result));
report(result);
Each run will generate a new secret point. When running this thousands of times it turns out that there is 1/9 probability that the algorithm needs a third guess. In the other 8/9 cases, the algorithm needs two guesses.
One idea is as follows:
You pick an initial random point, and for each dimension, find the exact value. How? For the sake of symmetry, suppose that you desire to find x of the target point. Increase by one the x, and compute the distance of the new point from the target point. If it goes further, it means that you should move in the opposite direction. Hence, you can run a binary search and get the distance to find the exact x of the target point. Otherwise, it means that you are going in the right direction along X-axis. So, do a binary search between all points with the same y and z such that their x values can change from x+1 to 100. A more formal solution comes in the following (just a pseudo-code).
You should also ask about the complexity of this solution. As the dimension of the point is constant (3) and checking these conditions take a constant time, the complexity of number of calling getDifference function is O(log(n)). What is n here? the length of valid range for coordinates (here is 100).
1. p: (x,y,z) <- Pick a random coordinate
2. dist: (dist_x, dist_y, dist_z) <- getDifference(p, TargetPoint)
3. For each dimension, do the following (i can be 0 (x), 1 (y), 2 (3)):
4. if(dist == 0):
5. isFound[i] <- p[i]
6. continue
7. new_i <- p[i] + 1
8. new_point <- p
9. new_point[i] <- new_i
10. new_dist <- getDifference(new_point, pointToFound)
11. if(new_dist == 0):
12. isFound[i] <- new_point[i];
13. continue
14. if(new_dist[i] > dist[i]):
15. isFound[i] <- binary_search for range [0, p[i]-1] to find the exact value of the pointToFound in dimension i
15. continue
16. else:
17. isFound[i] <- binary_search for range [p[i] + 1, 100] to find the exact value of the pointToFound in dimension i
18. continue
Following method will work for coordinates with positive or negative real values as well.
Let's say you are searching for the coordinates of point P. As the first query point, use origin O. Let the distance to the origin O be |PO|. At this point, you know that P is on the surface of sphere
(P.x)^2 + (P.y)^2 + (P.z)^2 = |PO|^2 (1)
As the second query point, use Q = (|PO|, 0, 0). Not likely but if you find the distance |PQ| zero, Q is the point you are looking for. Otherwise, you get another sphere equation, and you know that P is on the surface of this sphere as well:
(P.x - |PO|)^2 + (P.y)^2 + (P.z)^2 = |PQ|^2 (2)
Now, if you subtract (1) from (2), you get
(P.x - |PO|)^2 - (P.x)^2 = |PQ|^2 - |PO|^2 (3)
Since the only unknown in this equation is P.x you can get its value:
P.x = (((-|PQ|^2 + |PO|^2) / |PO|) + |PO|)/2)
Following similar steps, you can get P.y with R = (0, |PO|, 0) and P.z with S = (0, 0, |PO|). So, by using four query points O, Q, R, and S you can get the coordinates of P.
So I have some numbers x = 320232 y = 2301 z = 12020305. I want to round these numbers off using JavaScript so that they become x = 320000 y = 2300 z = 12000000.
I tried Math.round and Math.floor but turns out that they only work with decimal values like
a = 3.1; Math.round(a); // Outputs 3 and not whole numbers.
So my question is can we round of whole numbers using JavaScript and If yes then how?
Edit: I want it to the round of to the starting 3 digit places as seen in the variables above. Like If there was another variable called c = 423841 It should round off to become c = 424000.
You could work with the logarithm of ten and adjust the digits.
const
format = n => v => {
if (!v) return 0;
const l = Math.floor(Math.log10(Math.abs(v))) - n + 1;
return Math.round(v / 10 ** l) * 10 ** l;
};
console.log([0, -9876, 320232, 2301, 12020305, 123456789].map(format(3)));
The solution is to first calculate how many numbers need to be rounded away, and then use that in a round.
Math.round(1234/100)*100 would round to 1200 so we can use this to round. We then only need to determan what to replace 100 with in this example.
That is that would be a 1 followed by LENGTH - 3 zeros. That number can be calculated as it is 10 to the power of LENGTH - 3, in JS: 10 ** (length - 3).
var x = 320232;
var y = 2301;
var z = 12020305;
function my_round(number){
var org_number = number;
// calculate integer number
var count = 0;
if (number >= 1) ++count;
while (number / 10 >= 1) {
number /= 10;
++count;
}
// length - 3
count = Math.round(count) - 3;
if (count < 0){
count = 0;
}
// 10 to the power of (length - 3)
var helper = 10 ** count;
return Math.round(org_number/helper)*helper;
}
alert(my_round(x));
alert(my_round(y));
alert(my_round(z));
It is not the prettiest code, though I tried to make it explainable code.
This should work:
function roundToNthPlace(input, n) {
let powerOfTen = 10 ** n
return Math.round(input/powerOfTen) * powerOfTen;
}
console.log([320232, 2301,12020305, 423841].map(input => roundToNthPlace(input, 3)));
Output: [320000, 2000, 12020000, 424000]
This is part of a bigger problem I try to solve in an exercise. It looks like this:
x is 10 times more likely to appear than y.
z appears 2x less often than y.
I solved this by calculating a single unit like this:
const x = 100;
const y = 10;
const z = 5;
const unit = 100 / (x + y + z);
unit equals 0.87
So when I do (0.87) + (0.87 * 10) + (0.87 * 5) I get 100%(almost)
Then I generate a random number between 0 and 1.
const randomNumber = Math.random();
function getValue() {
if (randomNumber <= 0.87) {
console.log('x');
} else if (randomNumber > 0.87 && randomNumber < 95.7) {
console.log('y');
} else console.log('z');
}
getValue();
If value<0.87 then I log out x, if value < 0.87+(0.087*10) I log y etc
Can anyone recommend a more logical and elegant way than this?
Your way looks clean for me except the fact that randomNumber > 0.87 is redundant.
if you store the value x, y and z in an array, you can probably write some cleaner code for example:
let prob = [100, 10, 5];
let sum = prob.reduce((a, b) => a + b, 0);
let normalizedProb = prob.map(p => p / sum);
let cummulativeProb = normalizedProb.map((cummulative => p => cummulative += p)(0));
for (let i = 0; i <= 50; i++) {
let r = Math.random();
console.log(cummulativeProb.filter(p => r >= p).length);
}
Also, you may want to read this post for faster implementation (in python though). However, the code will be more complicated for sure.
Since the weights are small integers, you can duplicate the x, y and z in an array, and just pick one random cell of the array:
let choices = "zyyxxxxxxxxxxxxxxxxxxxx";
console.log(choices[Math.floor(Math.random() * 23)]);
Here the magic number 23 is the number of choices, 1+2+20; and Math.floor(Math.random() * 23) is a random integer uniformly at random in range [0, 22] (both bounds included). See also:
Generating random whole numbers in JavaScript in a specific range?
Is it possible to get a random number between 1-100 and keep the results mainly within the 40-60 range? I mean, it will go out of that range rarely, but I want it to be mainly within that range... Is it possible with JavaScript/jQuery?
Right now I'm just using the basic Math.random() * 100 + 1.
The simplest way would be to generate two random numbers from 0-50 and add them together.
This gives a distribution biased towards 50, in the same way rolling two dice biases towards 7.
In fact, by using a larger number of "dice" (as #Falco suggests), you can make a closer approximation to a bell-curve:
function weightedRandom(max, numDice) {
let num = 0;
for (let i = 0; i < numDice; i++) {
num += Math.random() * (max/numDice);
}
return num;
}
JSFiddle: http://jsfiddle.net/797qhcza/1/
You have some good answers here that give specific solutions; let me describe for you the general solution. The problem is:
I have a source of more-or-less uniformly distributed random numbers between 0 and 1.
I wish to produce a sequence of random numbers that follow a different distribution.
The general solution to this problem is to work out the quantile function of your desired distribution, and then apply the quantile function to the output of your uniform source.
The quantile function is the inverse of the integral of your desired distribution function. The distribution function is the function where the area under a portion of the curve is equal to the probability that the randomly-chosen item will be in that portion.
I give an example of how to do so here:
http://ericlippert.com/2012/02/21/generating-random-non-uniform-data/
The code in there is in C#, but the principles apply to any language; it should be straightforward to adapt the solution to JavaScript.
Taking arrays of numbers, etc. isn't efficient. You should take a mapping which takes a random number between 0 to 100 and maps to the distribution you need. So in your case, you could take f(x)=-(1/25)x2+4x to get a distribution with the most values in the middle of your range.
I might do something like setup a "chance" for the number to be allowed to go "out of bounds". In this example, a 20% chance the number will be 1-100, otherwise, 40-60:
$(function () {
$('button').click(function () {
var outOfBoundsChance = .2;
var num = 0;
if (Math.random() <= outOfBoundsChance) {
num = getRandomInt(1, 100);
} else {
num = getRandomInt(40, 60);
}
$('#out').text(num);
});
function getRandomInt(min, max) {
return Math.floor(Math.random() * (max - min + 1)) + min;
}
});
<script src="https://ajax.googleapis.com/ajax/libs/jquery/1.11.1/jquery.min.js"></script>
<button>Generate</button>
<div id="out"></div>
fiddle: http://jsfiddle.net/kbv39s9w/
I needed to solve this problem a few years ago and my solution was easier than any of the other answers.
I generated 3 randoms between the bounds and averaged them. This pulls the result towards the centre but leaves it completely possible to reach the extremities.
It looks stupid but you can use rand twice:
var choice = Math.random() * 3;
var result;
if (choice < 2){
result = Math.random() * 20 + 40; //you have 2/3 chance to go there
}
else {
result = Math.random() * 100 + 1;
}
Sure it is possible. Make a random 1-100. If the number is <30 then generate number in range 1-100 if not generate in range 40-60.
There is a lot of different ways to generate such random numbers. One way to do it is to compute the sum of multiple uniformly random numbers. How many random numbers you sum and what their range is will determine how the final distribution will look.
The more numbers you sum up, the more it will be biased towards the center. Using the sum of 1 random number was already proposed in your question, but as you notice is not biased towards the center of the range. Other answers have propose using the sum of 2 random numbers or the sum of 3 random numbers.
You can get even more bias towards the center of the range by taking the sum of more random numbers. At the extreme you could take the sum of 99 random numbers which each were either 0 or 1. That would be a binomial distribution. (Binomial distributions can in some sense be seen as the discrete version of normal distributions). This can still in theory cover the full range, but it has so much bias towards the center that you should never expect to see it reach the endpoints.
This approach means you can tweak just how much bias you want.
What about using something like this:
var loops = 10;
var tries = 10;
var div = $("#results").html(random());
function random() {
var values = "";
for(var i=0; i < loops; i++) {
var numTries = tries;
do {
var num = Math.floor((Math.random() * 100) + 1);
numTries--;
}
while((num < 40 || num >60) && numTries > 1)
values += num + "<br/>";
}
return values;
}
<script src="https://ajax.googleapis.com/ajax/libs/jquery/1.11.1/jquery.min.js"></script>
<div id="results"></div>
The way I've coded it allows you to set a couple of variables:
loops = number of results
tries = number of times the function will try to get a number between 40-60 before it stops running through the while loop
Added bonus: It uses do while!!! Awesomeness at its best
You can write a function that maps random values between [0, 1) to [1, 100] according to weight. Consider this example:
Here, the value 0.95 maps to value between [61, 100].
In fact we have .05 / .1 = 0.5, which, when mapped to [61, 100], yields 81.
Here is the function:
/*
* Function that returns a function that maps random number to value according to map of probability
*/
function createDistributionFunction(data) {
// cache data + some pre-calculations
var cache = [];
var i;
for (i = 0; i < data.length; i++) {
cache[i] = {};
cache[i].valueMin = data[i].values[0];
cache[i].valueMax = data[i].values[1];
cache[i].rangeMin = i === 0 ? 0 : cache[i - 1].rangeMax;
cache[i].rangeMax = cache[i].rangeMin + data[i].weight;
}
return function(random) {
var value;
for (i = 0; i < cache.length; i++) {
// this maps random number to the bracket and the value inside that bracket
if (cache[i].rangeMin <= random && random < cache[i].rangeMax) {
value = (random - cache[i].rangeMin) / (cache[i].rangeMax - cache[i].rangeMin);
value *= cache[i].valueMax - cache[i].valueMin + 1;
value += cache[i].valueMin;
return Math.floor(value);
}
}
};
}
/*
* Example usage
*/
var distributionFunction = createDistributionFunction([
{ weight: 0.1, values: [1, 40] },
{ weight: 0.8, values: [41, 60] },
{ weight: 0.1, values: [61, 100] }
]);
/*
* Test the example and draw results using Google charts API
*/
function testAndDrawResult() {
var counts = [];
var i;
var value;
// run the function in a loop and count the number of occurrences of each value
for (i = 0; i < 10000; i++) {
value = distributionFunction(Math.random());
counts[value] = (counts[value] || 0) + 1;
}
// convert results to datatable and display
var data = new google.visualization.DataTable();
data.addColumn("number", "Value");
data.addColumn("number", "Count");
for (value = 0; value < counts.length; value++) {
if (counts[value] !== undefined) {
data.addRow([value, counts[value]]);
}
}
var chart = new google.visualization.ColumnChart(document.getElementById("chart"));
chart.draw(data);
}
google.load("visualization", "1", { packages: ["corechart"] });
google.setOnLoadCallback(testAndDrawResult);
<script src="https://www.google.com/jsapi"></script>
<div id="chart"></div>
Here's a weighted solution at 3/4 40-60 and 1/4 outside that range.
function weighted() {
var w = 4;
// number 1 to w
var r = Math.floor(Math.random() * w) + 1;
if (r === 1) { // 1/w goes to outside 40-60
var n = Math.floor(Math.random() * 80) + 1;
if (n >= 40 && n <= 60) n += 40;
return n
}
// w-1/w goes to 40-60 range.
return Math.floor(Math.random() * 21) + 40;
}
function test() {
var counts = [];
for (var i = 0; i < 2000; i++) {
var n = weighted();
if (!counts[n]) counts[n] = 0;
counts[n] ++;
}
var output = document.getElementById('output');
var o = "";
for (var i = 1; i <= 100; i++) {
o += i + " - " + (counts[i] | 0) + "\n";
}
output.innerHTML = o;
}
test();
<pre id="output"></pre>
Ok, so I decided to add another answer because I felt like my last answer, as well as most answers here, use some sort of half-statistical way of obtaining a bell-curve type result return. The code I provide below works the same way as when you roll a dice. Therefore, it is hardest to get 1 or 99, but easiest to get 50.
var loops = 10; //Number of numbers generated
var min = 1,
max = 50;
var div = $("#results").html(random());
function random() {
var values = "";
for (var i = 0; i < loops; i++) {
var one = generate();
var two = generate();
var ans = one + two - 1;
var num = values += ans + "<br/>";
}
return values;
}
function generate() {
return Math.floor((Math.random() * (max - min + 1)) + min);
}
<script src="https://ajax.googleapis.com/ajax/libs/jquery/1.11.1/jquery.min.js"></script>
<div id="results"></div>
I'd recommend using the beta distribution to generate a number between 0-1, then scale it up. It's quite flexible and can create many different shapes of distributions.
Here's a quick and dirty sampler:
rbeta = function(alpha, beta) {
var a = 0
for(var i = 0; i < alpha; i++)
a -= Math.log(Math.random())
var b = 0
for(var i = 0; i < beta; i++)
b -= Math.log(Math.random())
return Math.ceil(100 * a / (a+b))
}
var randNum;
// generate random number from 1-5
var freq = Math.floor(Math.random() * (6 - 1) + 1);
// focus on 40-60 if the number is odd (1,3, or 5)
// this should happen %60 of the time
if (freq % 2){
randNum = Math.floor(Math.random() * (60 - 40) + 40);
}
else {
randNum = Math.floor(Math.random() * (100 - 1) + 1);
}
The best solution targeting this very problem is the one proposed by BlueRaja - Danny Pflughoeft but I think a somewhat faster and more general solution is also worth mentioning.
When I have to generate random numbers (strings, coordinate pairs, etc.) satisfying the two requirements of
The result set is quite small. (not larger than 16K numbers)
The result set is discreet. (like integer numbers only)
I usually start by creating an array of numbers (strings, coordinate pairs, etc.) fulfilling the requirement (In your case: an array of numbers containing the more probable ones multiple times.), then choose a random item of that array. This way, you only have to call the expensive random function once per item.
Distribution
5% for [ 0,39]
90% for [40,59]
5% for [60,99]
Solution
var f = Math.random();
if (f < 0.05) return random(0,39);
else if (f < 0.95) return random(40,59);
else return random(60,99);
Generic Solution
random_choose([series(0,39),series(40,59),series(60,99)],[0.05,0.90,0.05]);
function random_choose (collections,probabilities)
{
var acc = 0.00;
var r1 = Math.random();
var r2 = Math.random();
for (var i = 0; i < probabilities.length; i++)
{
acc += probabilities[i];
if (r1 < acc)
return collections[i][Math.floor(r2*collections[i].length)];
}
return (-1);
}
function series(min,max)
{
var i = min; var s = [];
while (s[s.length-1] < max) s[s.length]=i++;
return s;
}
You can use a helper random number to whether generate random numbers in 40-60 or 1-100:
// 90% of random numbers should be between 40 to 60.
var weight_percentage = 90;
var focuse_on_center = ( (Math.random() * 100) < weight_percentage );
if(focuse_on_center)
{
// generate a random number within the 40-60 range.
alert (40 + Math.random() * 20 + 1);
}
else
{
// generate a random number within the 1-100 range.
alert (Math.random() * 100 + 1);
}
If you can use the gaussian function, use it. This function returns normal number with average 0 and sigma 1.
95% of this number are within average +/- 2*sigma. Your average = 50, and sigma = 5 so
randomNumber = 50 + 5*gaussian()
The best way to do that is generating a random number that is distributed equally in a certain set of numbers, and then apply a projection function to the set between 0 and a 100 where the projection is more likely to hit the numbers you want.
Typically the mathematical way of achieving this is plotting a probability function of the numbers you want. We could use the bell curve, but let's for the sake of easier calculation just work with a flipped parabola.
Let's make a parabola such that its roots are at 0 and 100 without skewing it. We get the following equation:
f(x) = -(x-0)(x-100) = -x * (x-100) = -x^2 + 100x
Now, all the area under the curve between 0 and 100 is representative of our first set where we want the numbers generated. There, the generation is completely random. So, all we need to do is find the bounds of our first set.
The lower bound is, of course, 0. The upper bound is the integral of our function at 100, which is
F(x) = -x^3/3 + 50x^2
F(100) = 500,000/3 = 166,666.66666 (let's just use 166,666, because rounding up would make the target out of bounds)
So we know that we need to generate a number somewhere between 0 and 166,666. Then, we simply need to take that number and project it to our second set, which is between 0 and 100.
We know that the random number we generated is some integral of our parabola with an input x between 0 and 100. That means that we simply have to assume that the random number is the result of F(x), and solve for x.
In this case, F(x) is a cubic equation, and in the form F(x) = ax^3 + bx^2 + cx + d = 0, the following statements are true:
a = -1/3
b = 50
c = 0
d = -1 * (your random number)
Solving this for x yields you the actual random number your are looking for, which is guaranteed to be in the [0, 100] range and a much higher likelihood to be close to the center than the edges.
This answer is really good. But I would like to post implementation instructions (I'm not into JavaScript, so I hope you will understand) for different situation.
Assume you have ranges and weights for every range:
ranges - [1, 20], [21, 40], [41, 60], [61, 100]
weights - {1, 2, 100, 5}
Initial Static Information, could be cached:
Sum of all weights (108 in sample)
Range selection boundaries. It basically this formula: Boundary[n] = Boundary[n - 1] + weigh[n - 1] and Boundary[0] = 0. Sample has Boundary = {0, 1, 3, 103, 108}
Number generation:
Generate random number N from range [0, Sum of all weights).
for (i = 0; i < size(Boundary) && N > Boundary[i + 1]; ++i)
Take ith range and generate random number in that range.
Additional note for performance optimizations. Ranges don't have to be ordered neither ascending nor descending order, so for faster range look-up range that has highest weight should go first and one with lowest weight should go last.
I want to know if the JavaScript function Math.random uses a normal (vs. uniform) distribution or not.
If not, how can I get numbers which use a normal distribution? I haven't found a clear answer on the Internet, for an algorithm to create random normally-distributed numbers.
I want to rebuild a Schmidt-machine (German physicist). The machine produces random numbers of 0 or 1, and they have to be normally-distributed so that I can draw them as a Gaussian bell curve.
For example, the random function produces 120 numbers (0 or 1) and the average (mean) of these summed values has to be near 60.
Since this is the first Google result for "js gaussian random" in my experience, I feel an obligation to give an actual answer to that query.
The Box-Muller transform converts two independent uniform variates on (0, 1) into two standard Gaussian variates (mean 0, variance 1). This probably isn't very performant because of the sqrt, log, and cos calls, but this method is superior to the central limit theorem approaches (summing N uniform variates) because it doesn't restrict the output to the bounded range (-N/2, N/2). It's also really simple:
// Standard Normal variate using Box-Muller transform.
function gaussianRandom(mean=0, stdev=1) {
let u = 1 - Math.random(); //Converting [0,1) to (0,1)
let v = Math.random();
let z = Math.sqrt( -2.0 * Math.log( u ) ) * Math.cos( 2.0 * Math.PI * v );
// Transform to the desired mean and standard deviation:
return z * stdev + mean;
}
Normal Distribution Between 0 and 1
Building on Maxwell's Answer, this code uses the Box–Muller transform to give you a normal distribution between 0 and 1 inclusive. It just resamples the values if it's more than 3.6 standard deviations away (less than 0.02% chance).
function randn_bm() {
let u = 0, v = 0;
while(u === 0) u = Math.random(); //Converting [0,1) to (0,1)
while(v === 0) v = Math.random();
let num = Math.sqrt( -2.0 * Math.log( u ) ) * Math.cos( 2.0 * Math.PI * v );
num = num / 10.0 + 0.5; // Translate to 0 -> 1
if (num > 1 || num < 0) return randn_bm() // resample between 0 and 1
return num
}
Visualizations
n = 100
n = 10,000
n = 10,000,000
Normal Distribution With Min, Max, Skew
This version allows you to give a min, max, and skew factor. See my usage examples at the bottom.
function randn_bm(min, max, skew) {
let u = 0, v = 0;
while(u === 0) u = Math.random() //Converting [0,1) to (0,1)
while(v === 0) v = Math.random()
let num = Math.sqrt( -2.0 * Math.log( u ) ) * Math.cos( 2.0 * Math.PI * v )
num = num / 10.0 + 0.5 // Translate to 0 -> 1
if (num > 1 || num < 0)
num = randn_bm(min, max, skew) // resample between 0 and 1 if out of range
else{
num = Math.pow(num, skew) // Skew
num *= max - min // Stretch to fill range
num += min // offset to min
}
return num
}
randn_bm(-500, 1000, 1);
randn_bm(10, 20, 0.25);
randn_bm(10, 20, 3);
Here is the JSFiddle for these screenshots: https://jsfiddle.net/2uc346hp/
I want to know if the JavaScript function Math.random is normal distribution or not
Javascript Math.random is not a Normal Distribution(Gaussian bell curve). From ES 2015, 20.2.2.27 "Returns a Number value with positive sign, greater than or equal to 0 but less than 1, chosen randomly or pseudo randomly with approximately uniform distribution over that range, using an implementation-dependent algorithm or strategy. This function takes no arguments." So the provided collection when n is high enough we will get approximately uniform distribution. All values in the interval will have equal probability of appearance(straight line parallel to the x axis, denoting number between 0.0 and 1.0).
how can I get numbers which are normal distribution
There are several ways of getting collection of numbers with a normal distribution. As answered by Maxwell Collard the Box-Muller transform
does transform uniform distribution to normal distribution(the code can be found in Maxwell Collard answer).
An answer to another stackoverflow answer to a question has a reply with other uniform distribution to normal distribution algorithms. Such as:
Ziggurat,
Ratio-of-uniforms,
Inverting the CDF
Besides one of the answers says that: says:
The Ziggurat algorithm is pretty efficient for this, although the Box-Muller transform is easier to implement from scratch (and not crazy slow).
And finally
I want to rebuilt a Schmidt-machine (German physicist), the machine produces random numbers of 0 or 1 and they have to be normal distributed so I can draw them in Gaussian bell curve.
When we have only two values (0 or 1) Gaussian curve looks the same as uniform distribution with 2 possible values. That is why a simple
function randomZero_One(){
return Math.round(Math.random());
}
would suffice. It would return pseudo-randomly with approximately equal probability values 0 and 1.
I wanted to have approximately gaussian random numbers between 0 and 1, and after many tests I found this to be the best:
function gaussianRand() {
var rand = 0;
for (var i = 0; i < 6; i += 1) {
rand += Math.random();
}
return rand / 6;
}
And as a bonus:
function gaussianRandom(start, end) {
return Math.floor(start + gaussianRand() * (end - start + 1));
}
The Javascript Math.random() pseudorandom function returns variates that are equally distributed between 0 and 1. To get a Gaussian distribution I use this:
// returns a gaussian random function with the given mean and stdev.
function gaussian(mean, stdev) {
var y2;
var use_last = false;
return function() {
var y1;
if (use_last) {
y1 = y2;
use_last = false;
} else {
var x1, x2, w;
do {
x1 = 2.0 * Math.random() - 1.0;
x2 = 2.0 * Math.random() - 1.0;
w = x1 * x1 + x2 * x2;
} while (w >= 1.0);
w = Math.sqrt((-2.0 * Math.log(w)) / w);
y1 = x1 * w;
y2 = x2 * w;
use_last = true;
}
var retval = mean + stdev * y1;
if (retval > 0)
return retval;
return -retval;
}
}
// make a standard gaussian variable.
var standard = gaussian(100, 15);
// make a bunch of standard variates
for (i = 0; i < 1000; i++) {
console.log( standard() )
}
I think I got this from Knuth.
Plot can be seen here
Function that utilises the central limit theorem.
function normal(mu, sigma, nsamples){
if(!nsamples) nsamples = 6
if(!sigma) sigma = 1
if(!mu) mu=0
var run_total = 0
for(var i=0 ; i<nsamples ; i++){
run_total += Math.random()
}
return sigma*(run_total - nsamples/2)/(nsamples/2) + mu
}
From the spec:
15.8.2.14 random ( )
Returns a Number value with positive sign, greater than or equal to 0
but less than 1, chosen randomly or pseudo randomly with
approximately uniform distribution over that range, using an
implementation-dependent algorithm or strategy. This function takes no
arguments.
So, it's a uniform distribution, not normal or Gaussian. That's what you're going to find in just about any standard random number facility in any basic language runtime outside of specialized statistics libraries.
You are confusing the output of the function (which is a uniform distribution between 0 and 1) with the need to generate a Gaussian distribution by repeatedly drawing random numbers that are either 0 or 1 - after a large number of trials, their sum will be approximately normally distributed.
You can use the Math.random() function, then round the result to an integer: if it's < 0.5, return 0; if its >= 0.5, return 1. Now you have equal probabilities of zero and one, and you can continue with the approach you described in your question.
Just to clarify: I don't think it's possible to have an algorithm that produces either 0's or 1's in a normally distributed way - normal distribution requires a continuous variable.
When you do the above for say 120 numbers, you will on average get 60 1's and 60 0's. The actual distribution you get will be the binomial distribution with a mean of 60 and a standard deviation of
stdev = sqrt(p(1-p)N) = 5.48
The probability of a particular number k when you have n samples with probability p (which we fixed at 0.5) is
p = n! / ((n-k)! k!) p^k (1-p)^(n-k)
When p = 0.5, you end up with just the binomial coefficients - which approach the normal distribution for n > 30, typically.
And a single line example:
Math.sqrt(-2 * Math.log(Math.random()))*Math.cos((2*Math.PI) * Math.random())
and a Fiddle
https://jsfiddle.net/rszgjqf8/
For those interested in generating values of a normal distrubution, I would recommend checking this implementation of the Ziggurat algorithm in JavaScript: https://www.npmjs.com/package/node-ziggurat
The code of found in the author's page is:
function Ziggurat(){
var jsr = 123456789;
var wn = Array(128);
var fn = Array(128);
var kn = Array(128);
function RNOR(){
var hz = SHR3();
var iz = hz & 127;
return (Math.abs(hz) < kn[iz]) ? hz * wn[iz] : nfix(hz, iz);
}
this.nextGaussian = function(){
return RNOR();
}
function nfix(hz, iz){
var r = 3.442619855899;
var r1 = 1.0 / r;
var x;
var y;
while(true){
x = hz * wn[iz];
if( iz == 0 ){
x = (-Math.log(UNI()) * r1);
y = -Math.log(UNI());
while( y + y < x * x){
x = (-Math.log(UNI()) * r1);
y = -Math.log(UNI());
}
return ( hz > 0 ) ? r+x : -r-x;
}
if( fn[iz] + UNI() * (fn[iz-1] - fn[iz]) < Math.exp(-0.5 * x * x) ){
return x;
}
hz = SHR3();
iz = hz & 127;
if( Math.abs(hz) < kn[iz]){
return (hz * wn[iz]);
}
}
}
function SHR3(){
var jz = jsr;
var jzr = jsr;
jzr ^= (jzr << 13);
jzr ^= (jzr >>> 17);
jzr ^= (jzr << 5);
jsr = jzr;
return (jz+jzr) | 0;
}
function UNI(){
return 0.5 * (1 + SHR3() / -Math.pow(2,31));
}
function zigset(){
// seed generator based on current time
jsr ^= new Date().getTime();
var m1 = 2147483648.0;
var dn = 3.442619855899;
var tn = dn;
var vn = 9.91256303526217e-3;
var q = vn / Math.exp(-0.5 * dn * dn);
kn[0] = Math.floor((dn/q)*m1);
kn[1] = 0;
wn[0] = q / m1;
wn[127] = dn / m1;
fn[0] = 1.0;
fn[127] = Math.exp(-0.5 * dn * dn);
for(var i = 126; i >= 1; i--){
dn = Math.sqrt(-2.0 * Math.log( vn / dn + Math.exp( -0.5 * dn * dn)));
kn[i+1] = Math.floor((dn/tn)*m1);
tn = dn;
fn[i] = Math.exp(-0.5 * dn * dn);
wn[i] = dn / m1;
}
}
zigset();
}
Create a Ziggurat.js file and then:
var z = new Ziggurat();
z.nextGaussian();
For me it's working just perfect and as I had read in Wikipedia, this is a more efficient algorithm than the Box-Muller.
enter link description here
I have tested several functions with the right configuration all work similarly and well.
http://jsfiddle.net/p3y40gf3/29/
Central limit is nice, must be with (n=3 for 6) and 12 for 12 to look as others. I configured others also to (6) or 12 or 1/12 as standard deviation, not sure why 12.
Central limit is a tiny bit less centered than Box/Muller and Ziggurat.
Box/Muller and Ziggurat look exactly the same
this variant by Joe(https://stackoverflow.com/a/33567961/466363) does standard deviation correctly:
function normal(mu, sigma, nsamples){ // using central limit
if(!nsamples) nsamples = 3
if(!sigma) sigma = 1
if(!mu) mu=0
var run_total = 0
for(var i=0 ; i<nsamples ; i++){
run_total += Math.random()
}
return sigma*(run_total - nsamples/2)/(nsamples/2) + mu
}
Ziggurat is also nice but needs to be adjusted from z score to from 0 to 1 looks like it makes good numbers.
Box/Muller clipped is good but gives few repeated numbers at clipped edges
but it is very similar to others,
incorrect random numbers should be discarded not clipped.
function randn_bm() {
var u = 0, v = 0;
while(u === 0) u = Math.random(); //Converting [0,1) to (0,1)
while(v === 0) v = Math.random();
let num = Math.sqrt( -2.0 * Math.log( u ) ) * Math.cos( 2.0 * Math.PI * v );
num = num / 6.0 + 0.5; // Translate to 0 -> 1 // changed here 10 to 6
if(num>1||num<0) return randn_bm(); return num; // bad random numbers should be discared not clipped
//return Math.max(Math.min(num, 1), 0); // cap between 0 and 1
}
Central limit variant it is called Bates distribution that is average
https://en.wikipedia.org/wiki/Bates_distribution
not confused with Irwin Hall that is a sum
https://en.wikipedia.org/wiki/Irwin%E2%80%93Hall_distribution
https://en.wikipedia.org/wiki/Normal_distribution#Generating_values_from_normal_distribution
A non verbose function to sample a random value from a Gaussian distribution I wrote some time ago:
function gaussianRandom(mean, sigma) {
let u = Math.random()*0.682;
return ((u % 1e-8 > 5e-9 ? 1 : -1) * (Math.sqrt(-Math.log(Math.max(1e-9, u)))-0.618))*1.618 * sigma + mean;
}
It should work if you clamp the values to the range you want.
skewnormal from normal and normal01
skewnormal(min, max, ..) returns a random number from the normal distribution that has been streched and offset to range from min to max, exponentially skewed with skew, and truncated to sigma standard deviations (in reverse order). Broken up into logical steps normal and normal01 for clarity and to generate random numbers directly from these intermediate functions if desired. (Plus a bonus lognormal!)
/// skewnormal(..) returns a random number from the normal distribution that has
/// been streched and offset to range from `min` to `max`, skewed with `skew`,
/// and truncated to `sigma` standard deviations. See https://stackoverflow.com/a/74258559/213246
const skewnormal = (min, max, skew = 1, sigma = 8) => {
/// normal() returns a random number from the standard normal distribution.
/// Uses the Box-Muller transform.
const normal = () => Math.sqrt(-2.0 * Math.log(Math.random())) * Math.cos(2.0 * Math.PI * Math.random());
/// normal01(..) returns normally distributed random number, whose range is
/// truncated at `sigma` standard deviations and shifted to interval `[0, 1]`.
const normal01 = (sigma) => {
while (true) {
let num = normal() / (sigma + 0.0) + 0.5; // translate to [0, 1]
if (0 <= num && num <= 1) return num; // ok if in range, else resample
}
}
var num = normal01(sigma);
num = Math.pow(num, skew) // skew
num *= max - min // stretch to fill range
num += min // offset to min
return num;
}
/// lognormal() returns a random number from the log-normal distribution.
const lognormal = () => Math.exp(normal());
Based on another popular answer by joshuakcockrell. You may prefer this implementation because: 1. it's factored to portray intermediate functions, 2. it exposes mathematically relevant and useful sigma parameter, 3. it has better names and comments.
See the JSFiddle for the complete demo environment, which makes it easy to define then test and visualize your own random distribution functions as pictured below:
View interactive charts: https://jsfiddle.net/rgefzusq/34/show/ Playground: https://jsfiddle.net/rgefzusq/34/
This is my JavaScript implementation of Algorithm P (Polar method for normal deviates) from Section 3.4.1 of Donald Knuth's book The Art of Computer Programming:
function gaussian(mean, stddev) {
return function() {
var V1
var V2
var S
do{
var U1 = Math.random()
var U2 = Math.random()
V1 = 2*U1-1
V2 = 2*U2-1
S = V1*V1+V2*V2
}while(S >= 1)
if(S===0) return 0
return mean+stddev*(V1*Math.sqrt(-2*Math.log(S)/S))
}
}
Use it like that:
var standard_normal = gaussian(0,1)
var a_standard_normal_deviate = standard_normal()
I found this library that includes lots of useful Random functions. You can either install it via simjs from npm, or just take the random-node-*.js file out directly for what you need.
http://www.simjs.com/random.html
http://www.simjs.com/download.html
This is my solution to the problem, using the Marsaglia polar method. The range depends on the parameters you give, without parameters it almost never generates anything outside of the range.
As it generates two normally distributed numbers per iteration, I declared a variable under window.temp.spareNormal to grab the spare one if it's there. Might not be the best location for it, but hey.
You'd probably have to round the result in order to get what you want.
window.temp = {
spareNormal: undefined
};
Math.normal = function (mean, standardDeviation) {
let q, u, v, p;
mean = mean || 0.5;
standardDeviation = standardDeviation || 0.125;
if (typeof temp.spareNormal !== 'undefined') {
v = mean + standardDeviation * temp.spareNormal;
temp.spareNormal = undefined;
return v;
}
do {
u = 2.0 * Math.random() - 1.0;
v = 2.0 * Math.random() - 1.0;
q = u * u + v * v;
} while (q >= 1.0 || q === 0);
p = Math.sqrt(-2.0 * Math.log(q) / q);
temp.spareNormal = v * p;
return mean + standardDeviation * u * p;
}
for finding normal distribution of value:
getNormal = (x, mean, standardDeviation, ) => {
return (1 / standardDeviation * Math.sqrt(2 * (3, 14))) * Math.pow(Math.E, -Math.pow(x - mean, 2) / (2 * (standardDeviation * standardDeviation)));
}
The only sort of qualifications I have for this is having taken a single statistics class. If I get something wrong, please tell me, I'd like to learn more about statistics and I don't want to keep thinking something wrong.
If you want to create a random number generator that produces numbers in a normal distribution, you should be able to take samples from a uniform distribution, which is no problem. If you set up a basic random number generator that generates numbers in range a to b, the distribution of values produced will have µ = (a+b)/2 and σ = (b-a)/√12. If the mean of a a few sample of values (≥30) taken from this distribution is taken for many such samples, then for the sampling distribution µ (sample means) = µ (population mean) and σ (sample means' stdev) = σ (population stdev)/√n (number of values in the sample).
By controlling the mean and stdev of the original distribution, you can control the ending mean and standard deviation of a random number generator that produces a normal distribution.
function all_normal(mu, sigma, nsamp)
{
var total = 0;
for (var a = 0; a < nsamp; a ++)
{
total += rand_int(mu - (sigma * Math.sqrt(3 * nsamp)), mu + (sigma * Math.sqrt(3 * nsamp)));
}
return Math.ceil(total / nsamp);
}
Just in case: Math.pow(Math.random(), p).
For example:
function testR(max = 100, min = 0, p = 1, c = 20)
{
let t = [];
for (let i = 0; i < c; ++i)
{
t.push(Math.floor(Math.pow(Math.random(), p) * (max - min + 1) + min));
}
console.log(
`p = ${String(p).padStart(5)}`, '|',
t.sort(function (a, b) { return a - b; }).join(', ')
);
}
testR(9, 0, 10);
testR(9, 0, 2);
testR(9, 0, 1);
testR(9, 0, 0.5);
testR(9, 0, 0.1);
testR(9, 0, 0.05);
Results in client/JS console
jsFiddle graph test:
let iset = 0;
let gset;
function randn() {
let v1, v2, fac, rsq;
if (iset == 0) {
do {
v1 = 2.0*Math.random() - 1.0;
v2 = 2.0*Math.random() - 1.0;
rsq = v1*v1+v2*v2;
} while ((rsq >= 1.0) || (rsq == 0));
fac = Math.sqrt(-2.0*Math.log(rsq)/rsq);
gset = v1*fac;
iset = 1;
return v2*fac;
} else {
iset = 0;
return gset;
}
}
//This is what I use for a Normal-ish distribution random function.
function normal_random(){
var pos = [ Math.random(), Math.random() ];
while ( Math.sin( pos[0] * Math.PI ) > pos[1] ){
pos = [ Math.random(), Math.random() ];
}
return pos[0];
};
This function returns a value between 0 and 1. Values near 0.5 are returned most often.