make callback function takes original data array

[Xuefeng-Zhu]

https://developer.mozilla.org/en-US/docs/Web/JavaScript/Reference/Global_Objects/Array/map

Make each, map, reduce methods take callback function similar to the one in js Array.

[mateogianolio]

LGTM, but can I just request that you run benchmarks before and after to make sure the performance isn't affected by these changes?

I'm thinking it's slower to apply a callback function in every step of the loop as opposed to executing an expression, but I might be wrong.

I'm aware that this library needs browser benchmarks and benchmarks without BLAS enabled, but for now you can change benchmarks/vector.js to include ../vector instead of ../vectorious.

Hope this isn't too much trouble! I'm grateful for your interest in improving this library.

[Xuefeng-Zhu]

@mateogianolio

Here is the benchmark result for this PR

> vectorious@4.7.0 benchmark /Users/Frank/mygit/vectorious
> node ./benchmarks/vector.js && node ./benchmarks/matrix.js

a, b = Vector.random(1024)
Vector.zeros(1024) x 196,433 ops/sec ±3.21% (75 runs sampled)
Vector.ones(1024) x 191,771 ops/sec ±4.04% (66 runs sampled)
Vector.range(0, 1024) x 68,066 ops/sec ±2.76% (79 runs sampled)
Vector.combine(a, b) x 65,491 ops/sec ±6.38% (53 runs sampled)
Vector.add(a, b) x 184,138 ops/sec ±3.94% (59 runs sampled)
Vector.subtract(a, b) x 187,315 ops/sec ±4.04% (60 runs sampled)
Vector.scale(a, Math.random()) x 210,720 ops/sec ±1.60% (65 runs sampled)
Vector.project(a, b) x 194,279 ops/sec ±2.40% (68 runs sampled)
a.add(b) x 1,940,823 ops/sec ±1.58% (87 runs sampled)
a.subtract(b) x 1,973,579 ops/sec ±1.26% (87 runs sampled)
a.scale(Math.random()) x 5,110,356 ops/sec ±1.90% (85 runs sampled)
a.normalize() x 3,877,903 ops/sec ±2.25% (85 runs sampled)
a.dot(b) x 4,824,868 ops/sec ±1.03% (86 runs sampled)
a.magnitude() x 12,079,812 ops/sec ±1.32% (87 runs sampled)
a.angle(b) x 1,564,677 ops/sec ±0.71% (88 runs sampled)
a.project(b) x 1,836,948 ops/sec ±0.79% (91 runs sampled)
a, b = Matrix.random(128, 128)
Matrix.identity(128) x 11,587 ops/sec ±4.40% (76 runs sampled)
Matrix.magic(128) x 2,507 ops/sec ±3.25% (83 runs sampled)
Matrix.zeros(128, 128) x 11,120 ops/sec ±4.58% (71 runs sampled)
Matrix.ones(128, 128) x 11,396 ops/sec ±4.52% (72 runs sampled)
Matrix.plu(a) x 390 ops/sec ±1.85% (80 runs sampled)
Matrix.augment(a, b) x 3,329 ops/sec ±4.06% (77 runs sampled)
Matrix.add(a, b) x 9,808 ops/sec ±4.49% (77 runs sampled)
Matrix.subtract(a, b) x 9,931 ops/sec ±4.47% (78 runs sampled)
Matrix.scale(Math.random()) x 10,138 ops/sec ±4.59% (73 runs sampled)
Matrix.transpose(a):
a.add(b) x 137,383 ops/sec ±0.89% (89 runs sampled)
a.subtract(b) x 137,881 ops/sec ±0.59% (90 runs sampled)
a.scale(Math.random()) x 201,702 ops/sec ±0.81% (90 runs sampled)
a.multiply(b) x 5,346 ops/sec ±2.31% (80 runs sampled)
a.transpose() x 7,141 ops/sec ±4.07% (57 runs sampled)
a.gauss() x 6,430 ops/sec ±4.58% (79 runs sampled)
a.lu() x 590 ops/sec ±1.50% (83 runs sampled)
a.plu() x 764 ops/sec ±1.18% (84 runs sampled)
a.solve() x 719 ops/sec ±0.89% (87 runs sampled)
a.trace() x 420,398 ops/sec ±6.53% (67 runs sampled)
a.swap(i, j) x 106,601 ops/sec ±2.57% (85 runs sampled)

[Xuefeng-Zhu]

This is the bench mark for master

> vectorious@4.7.0 benchmark /Users/Frank/mygit/vectorious
> node ./benchmarks/vector.js && node ./benchmarks/matrix.js

a, b = Vector.random(1024)
Vector.zeros(1024) x 202,087 ops/sec ±3.01% (79 runs sampled)
Vector.ones(1024) x 190,404 ops/sec ±3.98% (63 runs sampled)
Vector.range(0, 1024) x 65,465 ops/sec ±3.11% (78 runs sampled)
Vector.combine(a, b) x 68,050 ops/sec ±5.71% (56 runs sampled)
Vector.add(a, b) x 195,546 ops/sec ±3.88% (63 runs sampled)
Vector.subtract(a, b) x 200,427 ops/sec ±3.52% (63 runs sampled)
Vector.scale(a, Math.random()) x 217,565 ops/sec ±1.13% (63 runs sampled)
Vector.project(a, b) x 193,003 ops/sec ±3.04% (64 runs sampled)
a.add(b) x 2,065,532 ops/sec ±0.88% (88 runs sampled)
a.subtract(b) x 2,075,626 ops/sec ±0.85% (88 runs sampled)
a.scale(Math.random()) x 5,479,279 ops/sec ±0.89% (89 runs sampled)
a.normalize() x 4,181,604 ops/sec ±0.96% (91 runs sampled)
a.dot(b) x 4,962,911 ops/sec ±1.02% (86 runs sampled)
a.magnitude() x 12,940,876 ops/sec ±0.92% (92 runs sampled)
a.angle(b) x 1,516,228 ops/sec ±1.08% (87 runs sampled)
a.project(b) x 1,814,127 ops/sec ±0.96% (86 runs sampled)
a, b = Matrix.random(128, 128)
Matrix.identity(128) x 11,132 ops/sec ±4.64% (70 runs sampled)
Matrix.magic(128) x 2,554 ops/sec ±1.38% (85 runs sampled)
Matrix.zeros(128, 128) x 11,493 ops/sec ±4.41% (73 runs sampled)
Matrix.ones(128, 128) x 10,975 ops/sec ±4.59% (71 runs sampled)
Matrix.plu(a) x 370 ops/sec ±2.25% (80 runs sampled)
Matrix.augment(a, b) x 3,093 ops/sec ±4.87% (62 runs sampled)
Matrix.add(a, b) x 7,906 ops/sec ±7.19% (64 runs sampled)
Matrix.subtract(a, b) x 9,108 ops/sec ±5.21% (73 runs sampled)
Matrix.scale(Math.random()) x 9,528 ops/sec ±5.37% (71 runs sampled)
Matrix.transpose(a):
a.add(b) x 126,750 ops/sec ±1.25% (84 runs sampled)
a.subtract(b) x 127,694 ops/sec ±1.02% (86 runs sampled)
a.scale(Math.random()) x 200,087 ops/sec ±0.74% (89 runs sampled)
a.multiply(b) x 4,596 ops/sec ±3.39% (70 runs sampled)
a.transpose() x 6,929 ops/sec ±4.92% (62 runs sampled)
a.gauss() x 6,591 ops/sec ±4.68% (78 runs sampled)
a.lu() x 585 ops/sec ±1.28% (82 runs sampled)
a.plu() x 746 ops/sec ±0.89% (83 runs sampled)
a.solve() x 668 ops/sec ±1.00% (83 runs sampled)
a.trace() x 385,709 ops/sec ±7.68% (67 runs sampled)
a.swap(i, j) x 98,739 ops/sec ±2.06% (81 runs sampled)