沿着3D JavaScript数组的单轴查找最小值/最大值

Question

I have a 3-dimensional array that contains sets of min/max latlng bounds for a collection of n polygons. I want to find the min and max coordinates from the set of all polygons.

My solution below works, but I find it ungainly. My real question is: Is there a way to obtain the min/max along an axis so that it returns an array of form [ [lat_min, lon_min], [lat_max, lon_max] ] without performing a reduce function on each point separately?

// Where bounds is an array of latlon bounds for n polygons:
// bounds = [
//   [ [min_lat_1, min_lon_1], [max_lat_1, max_lon_1] ],
//   ...
//   [ [min_lat_n, min_lon_n], [max_lat_n, max_lon_n] ]
// ]
const x1 = bounds.reduce((min, box) => {
  return box[0][0] < min ? box[0][0] : min;
}, bounds[0][0][0]);
const y1 = bounds.reduce((min, box) => {
  return box[0][1] < min ? box[0][1] : min;
}, bounds[0][0][1]);
const x2 = bounds.reduce((max, box) => {
  return box[1][0] > max ? box[1][0] : max;
}, bounds[0][1][0]);
const y2 = bounds.reduce((max, box) => {
  return box[1][1] > max ? box[1][1] : max;
}, bounds[0][1][1]);

EDIT: The responses I've gotten so far improve on my code, but so far nothing quite does what I'm hoping for.

Some further background/specification: I'm more familiar with python/numpy, where you can specify to apply a function across any axis. In this case, I would want to apply my function along axis 3, ie the depth axis. But, since I'm not just looking for min/max, the function I create would also need to return a function (Min or Max) depending on index. Is this simply not feasible in Javascript? It just seems like there should be some elegant combo in es6 that gets the job done.

Answer 1

You can use .apply on Math.min or Math.max to pass an array of numbers to compare at once.

You can also use .map to extract either X or Y of your coordinates.

 const bounds = [ [ [1, 2], [3, 999] ], [ [-999, 6], [7, 8] ], [ [9, 10], [11, 12] ] ] // `pos` determines whether it should get the min set or the max set const getXOf = pos => bounds.map(a => a[pos][0]); const getYOf = pos => bounds.map(a => a[pos][1]); // Using .apply, you can compare an array of numbers at once. const findMin = arr => Math.min.apply(null, arr); const findMax = arr => Math.max.apply(null, arr); // For clarity only const MIN = 0; const MAX = 1; const min_x = findMin(getXOf(MIN)); const min_y = findMin(getYOf(MIN)); const max_x = findMax(getXOf(MAX)); const max_y = findMax(getYOf(MAX)); console.log(min_x, min_y, max_x, max_y); 

Answer 2

You could combine all the reduces in just one and return the array [ [lat_min, lon_min], [lat_max, lon_max] ] like this:

 const bounds = [ [ [ 0.0, 0.1 ], [ 0.6, 0.7 ] ], [ [ 0.2, 0.3 ], [ 0.8, 0.9 ] ], [ [ 0.4, 0.5 ], [ 0.1, 0.2 ] ] ]; const minMax = bounds.reduce((current, box) => { const minLat = box[0][0] < current[0][0] ? box[0][0] : current[0][0]; const minLon = box[0][1] < current[0][1] ? box[0][1] : current[0][1]; const maxLat = box[1][0] > current[1][0] ? box[1][0] : current[1][0]; const maxLon = box[1][1] > current[1][1] ? box[1][1] : current[1][1]; return [ [ minLat, minLon ], [ maxLat, maxLon ] ] }, bounds[0]); console.log('combined reduce:', minMax); 

Below is your code for reference:

 const bounds = [ [ [ 0.0, 0.1 ], [ 0.6, 0.7 ] ], [ [ 0.2, 0.3 ], [ 0.8, 0.9 ] ], [ [ 0.4, 0.5 ], [ 0.1, 0.2 ] ] ]; const x1 = bounds.reduce((min, box) => { return box[0][0] < min ? box[0][0] : min; }, bounds[0][0][0]); const y1 = bounds.reduce((min, box) => { return box[0][1] < min ? box[0][1] : min; }, bounds[0][0][1]); const x2 = bounds.reduce((max, box) => { return box[1][0] > max ? box[1][0] : max; }, bounds[0][1][0]); const y2 = bounds.reduce((max, box) => { return box[1][1] > max ? box[1][1] : max; }, bounds[0][1][1]); console.log('separate reduce:', [ [ x1, y1 ], [ x2, y2 ] ]);