[Suggestion] Math.mean and Math.roundBy

[fega]

Hello, I really like this proposal, and I'd like two suggestions

Math.mean([1,2,3]) // 2

Math.roundBy(100.1234, 2) // 100.12

I'm proposing a separated method for round, since I think that Math.round could lead to breaking changes

[Crissov]

Means

I support the addition of a buildin function for the arithmetic mean, but I think mean() is not the best name for it, because there are other useful means as well, and even a generalized mean. Therefore, we should have all of them:

• mean(i[, x]*) = mean(i, args) generalized mean: 1/n(args)*pow(sum(args.map(x => pow(x, i))), 1/i)
• min(args) = mean(-∞, args) = quant(0, j, args) minimum
• hmean(args) = mean(-1, args) harmonic mean
• (ghmean(x, y, n) geometric-harmonic mean):
  for (var i = 1, g[0] = x, h[0] = y; i <= n && g[i] != h[i]; i++) {g[i] = gmean(g[i-1], h[i-1]); h[i] = hmean(g[i-1], h[i-1]);} return h[n];
• gmean(args) = mean(0, args) geometric mean, arithmetic-harmonic mean
• (agmean(x, y, n) arithmetic-geometric mean):
  for (var i = 1, a[0] = x, g[0] = y; i <= n && a[i] != g[i]; i++) {a[i] = amean(a[i-1], g[i-1]); g[i] = gmean(a[i-1], g[i-1]);} return g[n];
• amean(args) = avg(args) = mean(1, args) arithmetic mean, average
• qmean(args) = hypot(args) = mean(2, args) quadratic mean = root mean squared (RMS) = hypotenuse
• cmean(args) = mean(3, args) cubic mean
• max(args) = mean(+∞, args) = quant(n(args), j, args) maximum

Since you might want to support truncated means or weighted means out of the box as well, it might make sense to expect the arguments args as an array (or similar iteratable): mean(int = 1, args = [], trunc = 0, weight = []).

Other

Other functions working on a set of values would make sense as well:

• range(args) = max(args) - min(args)
• mid(args) = amean(max(args), min(args)) mid-range
• mode(args) mode, i.e. most frequent value
• median(args) = med(args) = quant(1, 2, args) = quart(2, args) median, i. e. central or middle value
• quant(i, j, args) quantile: j groups with equal number of members, i-th boundary; i.e. 0 ≤ i ≤ j
  • terc(i, args) = quant(i, 3, args) tercile
  • quart(i, args) = quant(i, 4, args) quartile
  • quint(i, args) = quant(i, 5, args) quintile
  • sext(i, args) = quant(i, 6, args) sextile
  • dec(i, args) = quant(i, 10, args) decile
  • perc(i, args) = quant(i, 100, args) percentile
• n(args) = count(args) (= args.length) number of arguments
• sum(args) = Σ(args) sum, i.e. result of adding all arguments Math.sum? #4
• product(args) = Π(args) product, i.e. result of multiplying all arguments

Rounding

For rounding #11, I would prefer a module as a second parameter that the result must be a multiple of: mround(1.2345, 0.01) = 1.23.

[Rudxain]

Also Standard Deviation may be a good idea