in my project I have to turn a bunch of coordinates to some meaningful two-dimensional array but I really don't know how to do it. Can somebody help?
To explain what I exactly want, let me give an example:
Let's suppose that I have these 2 arrays(the reason that I started from one is because 0 and the last element of my rows are borders):
[[1, 1], [1, 2], [1, 4], [1, 5], [1, 8], [1, 9], [1, 10],[2, 1], [2, 2], [2, 4], [2, 5], [2, 6], [2, 7], [2, 8], [2, 10]]
Let the value inside these coordinates be like [row,col]. And let's say I wan't to match them to generate some sort of two-dimensional array and each of the elements should contain the value '#'. However, for example;
[1, 2], [1, 4]
[2, 2], [2, 4]
If there's a coordinate missing between two of these elements, they should be separated, meaning that there should be two different two-dimensional arrays, being split from that coordinate. In this case, the result should be;
// First two-dimensional array
const firstArray = [
['#','#'],
['#','#']
]
const secondArray = [
['#','#','','','#','#','#'],
['#','#','#','#','#','','#'],
]
In the second array, there are some '' values, but that is because the there are some coordinates missing(for [1, 5] and [1, 8], [1,6] and [1,7] are missing). So that should be considered too.
If you didn't understand please comment under the question me so that I can explain it to you.
How can I come up with the functionality that I'm looking for?
You can accomplish both steps with a single Array#reduce() call by using the coordinates themselves to place each [row, col] in its relevant place in the matrix.
Here using an OR short circuit for assigning new sub-arrays, with a commented out replacement using the logical nullish assignment operator (??=), and the comma operator for shorthand return in the arrow function.
const coords = [[1, 1], [1, 2], [1, 4], [1, 5], [1, 8], [1, 9], [1, 10], [2, 1], [2, 2], [2, 4], [2, 5], [2, 6], [2, 7], [2, 8], [2, 10]];
const matrix = coords.reduce((acc, [row, col]) => (
// using OR short circuit for compatibility
(acc[row - 1] || (acc[row - 1] = []))[col - 1] = [row, col], acc
// using logical nullish assignment operator (??=)
//(acc[row - 1] ??= [])[col - 1] = [row, col], _matrix
), [])
// logging
for (const row of matrix) {
console.log(`[[${row.join('], [')}]]`)
}
.as-console-wrapper { max-height: 100% !important; top: 0; }
const input = [[1, 1], [1, 2], [1, 4], [1, 5], [1, 8], [1, 9], [1, 10],[2, 1], [2, 2], [2, 4], [2, 5], [2, 6], [2, 7], [2, 8], [2, 10]]
const result = input.reduce((acc, [x, y]) => {
acc[x - 1] ??= []
const previousY = acc[x - 1][acc[x-1].length - 1];
if (previousY) {
const delta = y - previousY;
if (delta > 1) acc[x-1].push(...Array.from({length: delta - 1}));
}
acc[x-1].push(y);
return acc
}, [])
console.log('1234567890')
console.log(
result.map(row =>
row.map(coor => coor ? '#' : ' ').join('')
).join('\n'))
Related
I have an object like this:
const arrays = {
one: [[1, 33, 41], [2, 0, 27], [3, 7, 9], [4, 1, 3]],
two: [[1, 77, 2], [2, 6, 3], [3, 0, 0], [4, 55, 3]],
three: [[1, 4, 6], [2, 0, 0], [3, 5, 6], [4, 0, 0]],
};
As you can see each first number is equal:
1 for each first inner array
2 for each second inner array
3 for each third inner array
etc...
I want to filter based on the first number of each array
and some comparator number e.g. [3]
If we have a filter number [3] (smaller or equal to 3),
the wanted result would be:
const arrays = {
one: [[1,33,41], [2,0,27], [3,7,9]],
two: [[1,77,2], [2,6,3], [3,0,0]],
three: [[1,4,6], [2,0,0], [3,5,6]],
};
Since all first numbers of inner arrays are smaller than or equal to 3.
The arrays starting with 4,5... are ignored.
What would be the ramda's way to have this functionality?
I like Ramda's map function because it can iterate over the properties of an object (and so avoid Object.fromEntries & Object.entries) and apply a function to each one of them. That function is filter which will take as argument the inner arrays. The function given to filter is itself a composition of gte and head; takes the first element of an array and compare it with 3:
const arrays =
{ one: [[1, 33, 41], [2, 0, 27], [3, 7, 9], [4, 1, 3]]
, two: [[1, 77, 2], [2, 6, 3], [3, 0, 0], [4, 55, 3]]
, three: [[1, 4, 6], [2, 0, 0], [3, 5, 6], [4, 0, 0]] };
map(filter(compose(gte(3), head)), arrays);
// ^ ^ ^ ^ ^
// A B C D E
//=> { one: [[ 1, 33, 41], [2, 0, 27], [3, 7, 9]]
//=> , two: [[ 1, 77, 2], [2, 6, 3], [3, 0, 0]]
//=> , three: [[ 1, 4, 6], [2, 0, 0], [3, 5, 6]] }
map over each property (A); each array is passed to filter (B)
Each inner array is passed to compose (C)
Take the head of each inner array (E) and compare with 3 (D)
Scott Christopher rightly pointed out in the comments that gte can be confusing when partially applied. In fact the whole composition can be replaced with this simple lambda: ([x]) => x <= 3.
Alternative solution which I like too:
map(filter(([x]) => x <= 3), arrays);
I'd totally subscribe for #customcommander's approach,
just wanted to add that you can also pass numerical indexes to R.propSatisfies.
const headIs3OrBelow = R.propSatisfies(R.gte(3), 0);
const fn = R.map(R.filter(headIs3OrBelow));
// ===
const data = {
one: [[1, 33, 41], [2, 0, 27], [3, 7, 9], [4, 1, 3]],
two: [[1, 77, 2], [2, 6, 3], [3, 0, 0], [4, 55, 3]],
three: [[1, 4, 6], [2, 0, 0], [3, 5, 6], [4, 0, 0]],
};
console.log(
fn(data),
);
<script src="https://cdnjs.cloudflare.com/ajax/libs/ramda/0.27.1/ramda.js" integrity="sha512-3sdB9mAxNh2MIo6YkY05uY1qjkywAlDfCf5u1cSotv6k9CZUSyHVf4BJSpTYgla+YHLaHG8LUpqV7MHctlYzlw==" crossorigin="anonymous" referrerpolicy="no-referrer"></script>
Also agree that gte and other similar methods are very difficult to read, because they kind of read backwards as is 3 gte than x... in Haskell you could do something like:
3 `gte` x
Vanilla approach:
const headIs3OrBelow = ([head]) => head <= 3;
const fn = (data) => Object.entries(data).reduce(
(res, [k, lists]) => ({ ...res, [k]: lists.filter(headIs3OrBelow) }),
{},
);
// ===
const data = {
one: [[1, 33, 41], [2, 0, 27], [3, 7, 9], [4, 1, 3]],
two: [[1, 77, 2], [2, 6, 3], [3, 0, 0], [4, 55, 3]],
three: [[1, 4, 6], [2, 0, 0], [3, 5, 6], [4, 0, 0]],
};
console.log(
fn(data),
);
I understand you would like to use Ramda, this is not a solution using the library but accomplishes the same. You could create an object from entries that are filtered by comparing the first array value a[0] to the maxVal passed to the function.
const arrays = {
one: [[1, 33, 41], [2, 0, 27], [3, 7, 9], [4, 1, 3]],
two: [[1, 77, 2], [2, 6, 3], [3, 0, 0], [4, 55, 3]],
three: [[1, 4, 6], [2, 0, 0], [3, 5, 6], [4, 0, 0]],
};
const filterArrays = (arsObj, maxVal) => {
return Object.fromEntries(Object.entries(arsObj).map(([k, v]) => {
return [k, v.filter((a) => a[0] <= maxVal)];
}));
}
const result = filterArrays(arrays, 3);
console.log(result);
I have been trying to accomplish this since yesterday, though no luck yet. I have found solutions where there always is a slight difference in what I want to accomplish.
I am trying to get all possible combinations, slightly like this: combination_k, but I also want the same items to pair up with itself, so given the following:
input [1, 4, 5] and 2 (number of combinations) should return:
[1, 1], [1, 4], [1, 5], [4, 4], [4, 5], [5, 5]
input [1, 4, 5] and 3 should return:
[1, 1, 1], [1, 1, 4], [1, 1, 5], [1, 4, 4], [1, 4, 5], [4, 4, 4], [4, 4, 5], [5, 5, 5], [5, 5, 4], [5, 5, 1] (The order is not important).
I have been adjusting combination_k, it got me far enough that it worked with 2 but it didn't work when I provided 3 as a parameter.
const combinations = getAllCombinations([1, 4, 5], 2);
// combinations = [1, 1], [1, 4], [1, 5], [4, 4], [4, 5], [5, 5]
Any tips are welcome!
The problem is commonly referred to as k-combinations with repetitions.
Here's a solution that relies on recursion to get the desired result:
const combinations = (array, r) => {
const result = [];
const fn = (array, selected, c, r, start, end) => {
if (c == r) {
result.push([...selected]);
return;
}
for (let i = start; i <= end; i++) {
selected[c] = array[i];
fn(array, selected, c + 1, r, i, end);
}
}
fn(array, [], 0, r, 0, array.length - 1);
return result;
}
console.log(combinations([1, 4, 5], 3));
A modified version of the code you provided:
function getAllCombinations(arr, n) {
if (n <= 0) return [];
if (n === 1) return [...arr];
return arr.reduce((acc, cur, i) => {
const head = arr.slice(i, i + 1);
const combinations = getAllCombinations(arr.slice(i), n - 1)
.map(x => head.concat(x));
return [...acc, ...combinations];
}, []);
}
console.log(getAllCombinations([1, 4, 5], 2).join('|'));
console.log(getAllCombinations([1, 4, 5], 3).join('|'));
If I have one array:
let x= [0,1,2,3,5]
And I have an array with several subarrays:
let winningIndices = [[0, 1, 2], [6, 7, 8], [0, 3, 6], [2, 5, 8], [0, 4, 8], [2, 4, 6]]
How can I check if array x contains all of the elements of any one subarray.
In other words, how can I check if array x has combinations of either the numbers 0,1,2 or 6,7,8...
Thanks in advance
"How can I check if array x contains all of the elements of any one subarray."
Here's the most straightforward functional interpretation.
const won = winningIndices.some(indices=>
indices.every(
item=>x.includes(item)
)
)
This could be an option. If you'd like to remove xIncludesItem key from result, just map it again.
const winningIndices = [
[0, 1, 2],
[6, 7, 8],
[0, 3, 6],
[2, 5, 8],
[0, 4, 8],
[2, 4, 6]
];
const x = [0, 1, 2, 3, 5];
const result = winningIndices
.map((item) => {
const xIncludesItem = item
.map((i) => x.includes(i))
.every(includes => includes);
return { item, xIncludesItem };
})
.filter(result => result.xIncludesItem);
console.log(result);
Array.includes is much slower than Set.has. A slightly more performant solution based on Ted Brownlow's solution would be:
const setX = new Set(x);
const won = winningIndices.some(indices=>
indices.every(
item=>setX.has(item)
)
)
I'm currently building a tic tac toe in vanilla javascript. However the game is 'sort of' done but I'm trying to add levels of difficulty. So basically the thing I want to do is , on every player move , to get the the closest possible winning combination based on his moves and place computer's mark into the missing winning's combinations place.
Let's say I have multidimensional array with the winning combinations
winningCombinations: [
[0, 1, 2],
[3, 4, 5],
[6, 7, 8],
[0, 4, 8],
[0, 3, 6],
[1, 4, 7],
[2, 4, 6],
[2, 5, 8]
]
And the player X starts his moves. So his first move is 0, so saving player's current combination in array. So in first move the current comb is
currentPlayerCombintation: [0]
so I want to return [0,1,2], [0,4,8] and [0,3,6] from the winning combination's array.
However the player makes his second move , so he target's 4 so
currentPlayerCombination: [0,4]
and now I want to return the closest possible winning combination which is [0,4,8].
I've tried a lot of things including every() , some() , filter() but could not achieve the thing I want.
I've tried sort of
for(let i = 0; i < this.currentPlayerCombination.length ; i++) {
this.winningCombinations.some((arr) => {
if(arr.includes(this.currentPlayerCombination[i])) {
console.log(arr);
}
});
}
But this didnt work as expected :(
You could take a Set and map the count of the matching items, get the max count and filter the array.
function getWinningPositions(pos) {
var posS = new Set(pos),
temp = winningCombinations.map(a => [a, a.reduce((c, v) => c + posS.has(v), 0)]),
max = Math.max(...temp.map(({ 1: c }) => c))
return temp
.filter(({ 1: c }) => c === max)
.map(([a]) => a);
}
var winningCombinations = [[0, 1, 2], [3, 4, 5], [6, 7, 8], [0, 4, 8], [0, 3, 6], [1, 4, 7], [2, 4, 6], [2, 5, 8]];
console.log(getWinningPositions([0]).map(a => a.join(' ')));
console.log(getWinningPositions([0, 4]).map(a => a.join(' ')));
console.log(getWinningPositions([0, 4, 5]).map(a => a.join(' ')));
.as-console-wrapper { max-height: 100% !important; top: 0; }
First map the winningCombinations to an array of arrays whose numbers are only the numbers that have not been picked yet. Then, find the lowest length of those arrays, and you can identify the original winningCombinations which are closest to the currentPlayerCombination:
const winningCombinations = [
[0, 1, 2],
[3, 4, 5],
[6, 7, 8],
[0, 4, 8],
[0, 3, 6],
[1, 4, 7],
[2, 4, 6],
[2, 5, 8]
];
const currentPlayerCombination = [0, 4];
// eg: get [[1, 2], [3, 5,], [6, 7, 8], [8], ...]
const winningCombsWithoutCurrent = winningCombinations.map(arr => (
arr.filter(num => !currentPlayerCombination.includes(num))
));
// eg: here, lowestLength should be 1, because [8] has a length of 1
const lowestLength = winningCombsWithoutCurrent.reduce((a, { length }) => Math.min(a, length), 3);
const combosWithLowestLength = winningCombsWithoutCurrent
.reduce((a, { length }, i) => {
if (length === lowestLength) {
a.push(winningCombinations[i]);
}
return a;
}, []);
console.log(combosWithLowestLength);
I have multiple arrays in a main/parent array like this:
var array = [[1, 17], [1, 17], [1, 17], [2, 12], [5, 9], [2, 12], [6, 2], [2, 12]];
here are the array's for simpler reading:
[1, 17]
[1, 17]
[1, 17]
[2, 12]
[5, 9]
[2, 12]
[6, 2]
[2, 12]
[2, 12]
I want to select the arrays that are repeated 3 or more times (> 3) and assign it to a variable. So in this example, var repeatedArrays would be [1, 17] and [2, 12].
So this should be the final result:
console.log(repeatedArrays);
>>> [[1, 17], [2, 12]]
I found something similar here but it uses underscore.js and lodash.
How could I it with javascript or even jquery (if need be)?
Try this
array.filter(( r={}, a=>!(2-(r[a]=++r[a]|0)) ))
var array = [[1, 17], [1, 17], [1, 17], [2, 12], [5, 9], [2, 12], [6, 2], [2, 12]];
var r= array.filter(( r={}, a=>!(2-(r[a]=++r[a]|0)) ))
console.log(JSON.stringify(r));
Time complexity O(n) (one array pass by filter function). Inspired by Nitish answer.
Explanation
The (r={}, a=>...) will return last expression after comma (which is a=>...) (e.g. (5,6)==6). In r={} we set once temporary object where we will store unique keys. In filter function a=>... in a we have current array element . In r[a] JS implicity cast a to string (e.g 1,17). Then in !(2-(r[a]=++r[a]|0)) we increase counter of occurrence element a and return true (as filter function value) if element a occurred 3 times. If r[a] is undefined the ++r[a] returns NaN, and further NaN|0=0 (also number|0=number). The r[a]= initialise first counter value, if we omit it the ++ will only set NaN to r[a] which is non-incrementable (so we need to put zero at init). If we remove 2- as result we get input array without duplicates - or alternatively we can also get this by a=>!(r[a]=a in r). If we change 2- to 1- we get array with duplicates only.
UPDATE
Even shorter version based on #ken comment can be written (it should always work with arrays of numbers). The original longer version of #ken code is in snippet and shows how #ken uses in clever way second argument of .filter to avoid usage global variable r.
array.filter(a=>!(2-(this[a]=++this[a]|0)))
var array = [[1, 17], [1, 17], [1, 17], [2, 12], [5, 9], [2, 12], [6, 2], [2, 12]];
var r= array.filter(a=>!(2-(this[a]=++this[a]|0)), {})
console.log(JSON.stringify(r));
You could take a Map with stringified arrays and count, then filter by count and restore the arrays.
var array = [[1, 17], [1, 17], [1, 17], [2, 12], [5, 9], [2, 12], [6, 2], [2, 12]],
result = Array
.from(array.reduce(
(map, array) =>
(json => map.set(json, (map.get(json) || 0) + 1))
(JSON.stringify(array)),
new Map
))
.filter(([, count]) => count > 2)
.map(([json]) => JSON.parse(json));
console.log(result);
.as-console-wrapper { max-height: 100% !important; top: 0; }
Filter with a map at wanted count.
var array = [[1, 17], [1, 17], [1, 17], [2, 12], [5, 9], [2, 12], [6, 2], [2, 12]],
result = array.filter(
(map => a =>
(json =>
(count => map.set(json, count) && !(2 - count))
(1 + map.get(json) || 1)
)
(JSON.stringify(a))
)
(new Map)
);
console.log(result);
.as-console-wrapper { max-height: 100% !important; top: 0; }
Unique!
var array = [[1, 17], [1, 17], [1, 17], [2, 12], [5, 9], [2, 12], [6, 2], [2, 12]],
result = array.filter(
(s => a => (j => !s.has(j) && s.add(j))(JSON.stringify(a)))
(new Set)
);
console.log(result);
.as-console-wrapper { max-height: 100% !important; top: 0; }
You can use Object.reduce, Object.entries for this like below
var array = [[1, 17], [1, 17], [1, 17], [2, 12], [5, 9], [2, 12], [6, 2], [2, 12]];
let res = Object.entries(
array.reduce((o, d) => {
let key = d.join('-')
o[key] = (o[key] || 0) + 1
return o
}, {}))
.flatMap(([k, v]) => v > 2 ? [k.split('-').map(Number)] : [])
console.log(res)
OR may be just with Array.filters
var array = [[1, 17], [1, 17], [1, 17], [1, 17], [2, 12], [5, 9], [2, 12], [6, 2], [2, 12]];
let temp = {}
let res = array.filter(d => {
let key = d.join('-')
temp[key] = (temp[key] || 0) + 1
return temp[key] == 3
})
console.log(res)
For a different take, you can first sort your list, then loop through once and pull out the elements that meet your requirement. This will probably be faster than stringifying keys from the array even with the sort:
var arr = [[1, 17], [1, 17], [1, 17], [2, 12], [5, 9], [2, 12], [6, 2], [2, 12]]
arr.sort((a, b) => a[0] - b[0] || a[1] - b[1])
// define equal for array
const equal = (arr1, arr2) => arr1.every((n, j) => n === arr2[j])
let GROUP_SIZE = 3
first = 0, last = 1, res = []
while(last < arr.length){
if (equal(arr[first], arr[last])) last++
else {
if (last - first >= GROUP_SIZE) res.push(arr[first])
first = last
}
}
if (last - first >= GROUP_SIZE) res.push(arr[first])
console.log(res)
You could also do this with a single Array.reduce where you would only push to a result property if the length is equal to 3:
var array = [[1, 17], [1, 17], [1, 17], [1, 17], [2, 12], [5, 9], [2, 12], [6, 2], [2, 12]];
console.log(array.reduce((r,c) => {
let key = c.join('-')
r[key] = (r[key] || 0) + 1
r[key] == 3 ? r.result.push(c) : 0 // if we have a hit push to result
return r
}, { result: []}).result) // print the result property
ES6:
const repeatMap = {}
array.forEach(arr => {
const key = JSON.stringify(arr)
if (repeatMap[key]) {
repeatMap[key]++
} else {
repeatMap[key] = 1
}
})
const repeatedArrays = Object.keys(repeatMap)
.filter(key => repeatMap[key] >= 3)
.map(key => JSON.parse(key))