We Keep Coding

Shuffle nested arrays in Javascript

I'm trying to sort multiple arrays within an array (which also has to be shuffled). A simplified example is:
let toShuffle = [
[1, 2, 3, 4, 5],
[9, 8, 7, 6, 5],
[10, 67, 19 ,27]
...
];
const shuffled = shuffle(toShuffle);
// outout would look something like:
// [
// [8, 6, 5, 7, 9],
// [4, 3, 1, 5, 2],
// [19, 26, 10, 67],
// ...
// ]
This needs to be flexible, so any number of arrays with any amount of values should be valid.
Here is what I've tried:
function shuffle(a) {
for (let e in a) {
if (Array.isArray(a[e])) {
a[e] = shuffle(a[e]);
} else {
a.splice(e, 1);
a.splice(Math.floor(Math.random() * a.length), 0, a[e]);
}
}
return a;
}
console.log("Shuffled: " + shuffle([
[1, 2, 3, 4, 5],
[5, 4, 3, 2, 1]
]))
But it's not working as intended. Is their an easier way to do this? Or is my code correct and just buggy.

You can use Array.from() to create a new shallow-copied array and then to shuffle Array.prototype.sort() combined with Math.random()
Code:
const toShuffle = [
[1, 2, 3, 4, 5],
[9, 8, 7, 6, 5],
[10, 67, 19 ,27]
]
const shuffle = a => Array.from(a).sort(() => .5 - Math.random())
const result = toShuffle.map(shuffle)
console.log('Shuffled:', JSON.stringify(result))
console.log('To shuffle:', JSON.stringify(toShuffle))

You almost got it. The problem is that you are removing one item from an array, instead of capturing the removed item and them placing in a random position:
let toShuffle = [
[1, 2, 3, 4, 5],
[9, 8, 7, 6, 5],
[10, 67, 19 ,27]
];
function shuffle(a) {
a = [...a]; //clone array
for (let e in a) {
if (Array.isArray(a[e])) {
a[e] = shuffle(a[e]);
} else {
a.splice(~~(Math.random() * a.length), 0, a.splice(e, 1)[0]);
}
}
return a;
}
console.log(JSON.stringify(shuffle(toShuffle)))
console.log(JSON.stringify(toShuffle))
[EDIT]
The original code did not shuffle the parent array, if you need shuffle everything recursively, you can use this:
let toShuffle = [
[1, 2, 3, 4, 5],
[9, 8, 7, 6, 5],
[10, 67, 19 ,27]
];
function shuffle(a) {
a = a.map(i => Array.isArray(i) ? shuffle(i) : i); //clone array
a.sort(i => ~~(Math.random() * 2) - 1); //shuffle
return a;
}
console.log("shuffled", JSON.stringify(shuffle(toShuffle)))
console.log("original", JSON.stringify(toShuffle))

Deleting arrays of same elements in 2 dimensional array in Javascript

I am wondering how you would go about deleting arrays that contain the same elements in a 2 dimensional array.
For example:
let 2dArr = [ [1, 2, 3],
[3, 2, 1],
[2, 4, 5],
[4, 5, 2],
[4, 3, 1] ];
This array would delete the second and fourth elements, returning the 2d array:
returnedArr = [ [1, 2, 3],
[2, 4, 5],
[4, 3, 1] ];
How exactly could this be done, preserving the 2d array? I could only think to loop through elements, comparing elements via a sort, and deleting them as you go along, but this would result in an indexing error if an element is deleted.

1) You can easily achieve the result using reduce and Set as:
let twodArr = [
[1, 2, 3],
[3, 2, 1],
[2, 4, 5],
[4, 5, 2],
[4, 3, 1],
];
const set = new Set();
const result = twodArr.reduce((acc, curr) => {
const key = [...curr].sort((a, b) => a - b).join();
if (!set.has(key)) {
set.add(key);
acc.push(curr);
}
return acc;
}, []);
console.log(result);
2) You can also use filter as:
let twodArr = [
[1, 2, 3],
[3, 2, 1],
[2, 4, 5],
[4, 5, 2],
[4, 3, 1],
];
const set = new Set();
const result = twodArr.filter((curr) => {
const key = [...curr].sort((a, b) => a - b).join();
return !set.has(key) ? (set.add(key), true) : false;
});
console.log(result);

const seen = []
const res = array.filter((item) => {
let key = item.sort().join()
if(!seen.includes(key)){
seen.push(key)
return item
}
})
console.log(res)

You can use hash map
let arr = [ [1, 2, 3], [3, 2, 1],[2, 4, 5],[4, 5, 2],[4, 3, 1] ];
let obj = {}
let final = []
for(let i=0; i<arr.length; i++){
// create a key
let sorted = [...arr[i]].sort((a,b)=> a- b).join`,`
// check if this is not present in our hash map
// add value to final out and update hash map accordingly
if(!obj[sorted]){
obj[sorted] = true
final.push(arr[i])
}
}
console.log(final)

Using Array.prototype.filter() and a Set as thisArg
let arr = [ [1, 2, 3],
[3, 2, 1],
[2, 4, 5],
[4, 5, 2],
[4, 3, 1] ];
let res = arr.filter(function(e){
const sorted = [...e].sort((a,b) => a-b).join('|');
return this.has(sorted) ? false : this.add(sorted)
},new Set)
console.log(res)

picking a unique set from an array of arrays

I attempted to ask a more complicated of this before but I couldn't explain it well so I am trying again with a simplified use case.
I will have an array of arrays like the following
var allData = [[1,2,3,4,5],[1,2,3,4,5],[1,2,3,4,5],[1,2,3,4,5],[1,2,3,4,5]]
I need to select 1 element from each array so that I get a unique set like [2,4,1,3,5] easy to do in this case as each array has all values. However this will rarely be the case. Instead I may have
var allData = [[1,2,4],[1,2],[1,2],[2,4,5],[1,2,3,5]]
In this case I couldn't pick 1 or 2 from the first array as that would prevent the 2nd and 3rd from having a unique combination. So something like [4,2,1,5,3] or [4,1,2,5,3] would be the only two possible answers for this combination.
The only way I see to do this is to just go through every combination but these will get fairly large so it doesn't seem reasonable as this happens real time. There are going to be at least 7 arrays, possibly 14 and distantly possible to have 31 so going through every combination would be fairly rough.
The 2nd part is if there is some way to "know" you have the best possible option. Say if there was some way I would know that having a single duplicate is my best case scenario. Even if I have to brute force it if I encounter a 1 duplication solution I would know to stop.
One easy way to get a very simple of this is to just subtract the number of possible choices from the number of elements but this is the correct answer in only the simplest of cases. Is there some type of library or anything to help solve these types of problems? It is a bit beyond my math abilities.
Here is something I have tried but it is too slow for larger sets and can fail. It works sometimes for the 2nd case I presented but only on luck
const allData = [[1,2,4],[1,2],[1,2],[2,4,5],[1,2,3,5]]
var selectedData = []
for (var i in allData){
console.log("length",allData[i].length)
var j = 0
while(j < allData[i].length){
console.log("chekcing",allData[i][j])
if (selectedData.includes(allData[i][j])){
console.log("removing item")
allData[i].splice(j,1)
}
else{j++}
}
var uniqueIds = Object.keys(allData[i])
console.log(uniqueIds)
var randId = Math.floor(Math.random() * uniqueIds.length)
console.log(randId)
selectedData.push(allData[i][randId])
console.log("selectedData",selectedData)
}

You can start with a fairly simple backtracking algorithm:
function pick(bins, n = 0, res = {}) {
if (n === bins.length) {
return res
}
for (let x of bins[n]) {
if (!res[x]) {
res[x] = n + 1
let found = pick(bins, n + 1, res)
if (found)
return found
res[x] = 0
}
}
}
//
let a = [[1, 2, 4], [1, 2], [1, 2], [2, 4, 5], [1, 2, 3, 4]]
console.log(pick(a))
This returns a mapping item => bin index + 1, which is easy to convert back to an array if needed.
This should perform relatively well for N < 10, for more/larger bins you can think of some optimizations, for example, avoid the worst case scenario by sorting bins from smallest to longest, or, depending on the nature of elements, represent bins as bitmasks.

You could count all elements and take various comparison with same indices.
function x([...data]) {
while (data.some(Array.isArray)) {
const
counts = data.reduce((r, a, i) => {
if (Array.isArray(a)) a.forEach(v => (r[JSON.stringify(v)] = r[JSON.stringify(v)] || []).push(i));
return r;
}, {}),
entries = Object.entries(counts),
update = ([k, v]) => {
if (v.length === 1) {
data[v[0]] = JSON.parse(k);
return true;
}
};
if (entries.some(update)) continue;
const grouped = entries.reduce((r, [, a]) => {
const key = JSON.stringify(a);
r[key] = (r[key] || 0) + 1;
return r;
}, {});
Object.entries(grouped).forEach(([json, length]) => {
const indices = JSON.parse(json);
if (indices.length === length) {
let j = 0;
indices.forEach(i => data[i] = data[i][j++]);
return;
}
if (length === 1) {
const value = JSON.parse(entries.find(([_, a]) => JSON.stringify(a) === json)[0]);
indices.forEach(i => data[i] = data[i].filter(v => v !== value));
data[indices[0]] = value;
}
});
}
return data;
}
console.log(...x([[1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5]]));
console.log(...x([[1, 2, 4], [1, 2], [1, 2], [2, 4, 5], [1, 2, 3, 5]]));
console.log(...x([[1, 2, 4], [1, 2], [1, 2], [2, 4, 5], [1, 2, 3, 5], [6, 7, 8, 9], [6, 7, 8, 9], [6, 7, 8, 10], [6, 7, 8, 10], [6, 7, 8, 10]]));

Here is an implementation based around counting occurrences across the arrays.
It first creates a map indexed by value counting the number of inner arrays each value occurs in. It then sorts by inner array length to prioritize shorter arrays, and then iterates over each inner array, sorting by occurrence and selecting the first non-duplicate with the lowest count, or, if there are no unique values, the element with the lowest count.
const
occurrencesAcrossArrays = (arr) =>
arr
.reduce((a, _arr) => {
[...new Set(_arr)].forEach(n => {
a[n] = a[n] || 0;
a[n] += 1;
});
return a;
}, {}),
generateCombination = (arr) => {
const dist = occurrencesAcrossArrays(arr)
return arr
.sort((a, b) => a.length - b.length)
.reduce((a, _arr) => {
_arr.sort((a, b) => dist[a] - dist[b]);
let m = _arr.find(n => !a.includes(n));
if (m !== undefined) {
a.push(m);
} else {
a.push(_arr[0]);
}
return a;
}, []);
};
console.log(generateCombination([[1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5]]).toString());
console.log(generateCombination([[1, 2, 4], [1, 2], [1], [2, 4, 5], [1, 2, 3, 5]]).toString());
console.log(generateCombination([[1, 2, 4], [1, 2], [1, 2], [2, 4, 5], [1, 2, 3, 5], [6, 7, 8, 9], [6, 7, 8, 9], [6, 7, 8, 10], [6, 7, 8, 10], [6, 7, 8, 10]]).toString());
Edit
In response to your comment – The situation seems to be emerging because the values all have the same occurrence count and are sequential.
This can be solved by keeping a running count of each value in the result array, and sorting each inner array by both by this running occurrence count as well as the original distribution count.This adds complexity to the sort, but allows you to simply access the first element in the array (the element with the lowest rate of occurrence in the result with the lowest occurrence count across all arrays).
const
occurrencesAcrossArrays = (arr) =>
arr
.reduce((a, _arr) => {
[...new Set(_arr)].forEach(n => {
a[n] = a[n] || 0;
a[n] += 1;
});
return a;
}, {}),
generateCombination = (arr) => {
const dist = occurrencesAcrossArrays(arr)
return arr
.sort((a, b) => a.length - b.length)
.reduce((acc, _arr) => {
_arr.sort((a, b) => (acc.occurrences[a] || 0) - (acc.occurrences[b] || 0) || dist[a] - dist[b]);
let m = _arr[0]
acc.occurrences[m] = acc.occurrences[m] || 0;
acc.occurrences[m] += 1;
acc.result.push(m);
return acc;
}, { result: [], occurrences: {} })
.result; // return the .result property of the accumulator
};
console.log(generateCombination([[2, 3, 4, 5, 6], [2, 3, 4, 5, 6], [2, 3, 4, 5, 6], [2, 3, 4, 5, 6], [2, 3, 4, 5, 6], [2, 3, 4, 5, 6], [2, 3, 4, 5, 6]]).toString());
// 2,3,4,5,6,2,3
console.log(generateCombination([[1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5]]).toString());
// 1,2,3,4,5
console.log(generateCombination([[1, 2, 4], [1, 2], [1], [2, 4, 5], [1, 2, 3, 5]]).toString());
// 1,2,4,5,3
console.log(generateCombination([[1, 2, 4], [1, 2], [1, 2], [2, 4, 5], [1, 2, 3, 5], [6, 7, 8, 9], [6, 7, 8, 9], [6, 7, 8, 10], [6, 7, 8, 10], [6, 7, 8, 10]]).toString());
//1,2,4,5,3,9,6,10,7,8
console.log(generateCombination([[1], [2, 3,], [3, 4, 5], [3, 4, 5, 6], [2, 3, 4, 5, 6, 7]]).toString());
// 1,2,4,6,7
A note on .reduce()
If you're having trouble getting your head around .reduce() you can rewrite all the instances of it in this example using .forEach() and declaring accumulator variables outside of the loop. (This will not always be the case, depending on how you manipulate the accumulator value within a reduce() call).
Example below:
const occurrencesAcrossArrays = (arr) => {
const occurrences = {};
arr.forEach(_arr => {
[...new Set(_arr)].forEach(n => {
occurrences[n] = occurrences[n] || 0;
occurrences[n] += 1;
});
});
return occurrences;
};
const generateCombination = (arr) => {
const dist = occurrencesAcrossArrays(arr);
const result = [];
const occurrences = {};
arr.sort((a, b) => a.length - b.length);
arr.forEach(_arr => {
_arr.sort((a, b) => (occurrences[a] || 0) - (occurrences[b] || 0) || dist[a] - dist[b]);
let m = _arr[0]
occurrences[m] = occurrences[m] || 0;
occurrences[m] += 1;
result.push(m);
});
return result;
};
console.log(generateCombination([[2, 3, 4, 5, 6], [2, 3, 4, 5, 6], [2, 3, 4, 5, 6], [2, 3, 4, 5, 6], [2, 3, 4, 5, 6], [2, 3, 4, 5, 6], [2, 3, 4, 5, 6]]).toString());
// 2,3,4,5,6,2,3
console.log(generateCombination([[1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5], [1, 2, 3, 4, 5]]).toString());
// 1,2,3,4,5
console.log(generateCombination([[1, 2, 4], [1, 2], [1], [2, 4, 5], [1, 2, 3, 5]]).toString());
// 1,2,4,5,3
console.log(generateCombination([[1, 2, 4], [1, 2], [1, 2], [2, 4, 5], [1, 2, 3, 5], [6, 7, 8, 9], [6, 7, 8, 9], [6, 7, 8, 10], [6, 7, 8, 10], [6, 7, 8, 10]]).toString());
//1,2,4,5,3,9,6,10,7,8
console.log(generateCombination([[1], [2, 3,], [3, 4, 5], [3, 4, 5, 6], [2, 3, 4, 5, 6, 7]]).toString());
// 1,2,4,6,7

You could solve this problem using a MILP-model. Here is one implementation in MiniZinc (data has been extended to seven days):
int: Days = 7;
int: Items = 5;
set of int: DAY = 1..Days;
set of int: ITEM = 1..Items;
array[DAY, ITEM] of 0..1: A = % whether item k is allowed on day i
[| 1, 1, 0, 1, 0
| 1, 1, 0, 0, 0
| 1, 1, 0, 0, 0
| 0, 1, 0, 1, 1
| 1, 1, 0, 0, 0
| 0, 1, 0, 1, 1
| 1, 1, 1, 0, 1 |];
array[DAY, ITEM] of var 0..1: x; % 1 if item selected k on day i, otherwise 0
array[DAY, DAY, ITEM] of var 0..1: w; % 1 if item k selected on both day i and day j, otherwise 0
% exactly one item per day
constraint forall(i in DAY)
(sum(k in ITEM)(x[i, k]) = 1);
% linking variables x and w
constraint forall(i, j in DAY, k in ITEM where i < j)
(w[i, j, k] <= x[i, k] /\ w[i, j, k] <= x[j, k] /\ w[i, j, k] >= x[i, k] + x[j, k] - 1);
% try to minimize duplicates and if there are duplicates put them as far apart as possible
var int: obj = sum(i, j in DAY, k in ITEM where i < j)(((Days - (j - i))^2)*w[i, j, k]);
solve minimize obj;
output
["obj="] ++ [show(obj)] ++
["\nitem="] ++ [show([sum(k in ITEM)(k*x[i, k]) | i in DAY])];
Running gives:
obj=8
item=[2, 1, 5, 4, 3, 2, 1]
The following package looks promising for a JavaScript implementation: https://www.npmjs.com/package/javascript-lp-solver

Javascript ES6 recursion function

I wonder how does this recursion function return numbers in this empty array(return n < 1 ? []) without pushing from [n, ...countdown(n - 1)], or how does it push? maybe I don't understand that es6 syntax
function countdown(n){ return n < 1 ? [] : [n, ...countdown(n - 1)] }
console.log(countdown(5)) log: [5, 4, 3, 2, 1]

The call stack steps that produce the result:
1. n = 5 => [5, ...countdown(4)]
2. n = 4 => [5, 4, ...countdown(3)]]
3. n = 3 => [5, 4, 3, ...countdown(2)]]
4. n = 2 => [5, 4, 3, 2, ...countdown(1)]
5. n = 1 => [5, 4, 3, 2, 1, ...countdown(0)]
6. n = 0 => [5, 4, 3, 2, 1, ...[]]
So the ... is spreading syntax notation that works like concat for two arrays.
Another schema could be like:
1. n = 5 => [5, ...countdown(4)]
2. n = 4 => [5, ...[4, ...countdown(3)]]
3. n = 3 => [5, ...[4, ...[3, ...countdown(2)]]]
4. n = 2 => [5, ...[4, ...[3, ..[2, ...countdown(1)]]]]
5. n = 1 => [5, ...[4, ...[3, ...[2, ...[1, ...countdown(0)]]]]]
6. n = 0 => [5, ...[4, ...[3, ...[2, ...[1, ...[]]]]]]
Think about that like function call stack from left to right:
x1( x2( x3() ) ) => ...( ...( ...(n) ) )

The ...-prefix syntax in this specific context (called the "spread syntax") puts the elements of the ...-prefixed array directly into the surrounding array.
[ 5, ...[4, 3, 2, 1] ]
is the same as
[5, 4, 3, 2, 1]
That's just what is happening here. The return values of countdown (from base case up to countdown(5)) look like:
[]
[1, ...[]] => [1]
[2, ...[1]] => [2, 1]
[3, ...[2, 1]] => [3, 2, 1]
[4, ...[3, 2, 1]] => [4, 3, 2, 1]
[5, ...[4, 3, 2, 1]] => [5, 4, 3, 2, 1]

How to make _lodash.zip with less code

Definition：Creates an array of grouped elements, the first of which contains the first elements of the given arrays, the second of which contains the second elements of the given arrays, and so on.
Current Solution：
const zip = (...arr) => {
let maxLength = 0
let res = []
for (let el of arr) {
maxLength = Math.max(maxLength, el.length)
}
for (let j = 0; j < maxLength; j++) {
const foo = []
for (let n of arr) {
foo.push(n[j])
}
res.push(foo)
}
return res
}
Test Case:
test(('zip', () => {
expect(zip([1, 2], [4, 5], [9, 1])).toEqual([[1, 4, 9], [2, 5, 1]])
}
test('zip', () => {
expect(zip([1, 2, 3], [4, 5, 6])).toEqual([[1, 4], [2, 5], [3, 6]])
})
test('zip', () => {
expect(zip([1, 2], [], [3, 4, 5])).toEqual([
[1, undefined, 3],
[2, undefined, 4],
[undefined, undefined, 5],
])
})
I want to get a better way to achieve zip, current solution is ugly

See Destructuring Assignment and Array.prototype.map for more info.
// Proof.
const zip = (...args) => [...new Array(Math.max(...args.map(arr => arr.length)))].map((x, i) => args.map((y) => y[i]))
// Proof.
console.log(zip([1, 2], [4, 5], [9, 1])) // [[1, 4, 9], [2, 5, 1]]
console.log(zip([1, 2, 3], [4, 5, 6])) // [[1, 4], [2, 5], [3, 6]]
console.log(zip([1, 2], [], [3, 4, 5])) // [[1, undefined, 3], [2, undefined, 4], [undefined, undefined, 5]]