random-gauge-value.js

/* files: ../point_src/core/head.js ../point_src/point-content.js pointlist point stage mouse dragging stroke ../point_src/functions/clamp.js ../point_src/text/beta.js ../point_src/functions/rel.js --- The _red_ gauge presents a raw `Math.random()`. The _green_ gauge presents a random with a calculated bias, to correct the long-term drift of a random function. Over time you'll see the unbiased gauge slowly drift away from a zero summation, where the biased gauge always tends towards zero. */ class MainStage extends Stage { canvas = 'playspace' /* Update speed of 1 is fastest. */ updateSpeed = 1 mounted(){ let a = this.a = (new Point(this.center)).update({radius: 200, y: rel(-50)}) let b = this.b = this.a.copy().update({y: rel(100)}) this.modu = 0 b.total = a.total = 0 /* The Bias pushes the _needle_ in a preferred direction. In theory .5 is _no bias_ But the -1 to 1 may natually walk away from the bias. On _my machine_ this bias tends to negative at `.4999`, and tends to positive at `.49999`. So a heustic _nudge_ may be prudent. */ b.bias = a.bias = .5 b.biasOffset = a.biasOffset = .00001 this.dragging.add(this.a, this.b) console.log('Mount') addControl('updateSpeed', { field: 'range' , label: 'update speed' , step: 1 , max: 200 , stage: this , onchange(ev) { /*slider changed. */ // debugger; let sval = ev.currentTarget.value this.stage.updateSpeed = parseInt(Math.sqrt(sval)*2) } }) } draw(ctx) { this.clear(ctx) this.modu += 1 this.modu % this.updateSpeed == 0 && this.updateWalkers(1, .5) // this.a.pen.fill(ctx, '#222255') ctx.fillStyle = '#eeddcc' ctx.font = 'normal 20px Courier New' let b = this.b; b.pen.indicator(ctx) let t1 = `${b.total.toFixed(1)}\n${b.biasOffset.toFixed(3)}` b.text.string(ctx, t1) let a = this.a; a.pen.indicator(ctx, {color: 'red'}) let t = `${a.total.toFixed(1)}\n${a.biasOffset.toFixed(3)}` a.text.string(ctx, t) } roll(p){ return (-1 + Math.random()) + (.5 + Math.random()) - p.bias - p.biasOffset } lesserRoll(){ /* The more _pure_ form of random tends to drift into positive. */ return -.5 + Math.random() } updateWalkers(iterationLimit=1, max=1) { let correction = .00001 let a = this.a let b = this.b let r = this.lesserRoll() a.total += r a.biasOffset += a.total * correction a.rotation += (1 * r) let r2 = this.roll(b) b.total += r2 b.biasOffset += (b.total * correction) * (Math.sqrt(correction ?? 1)*1000) b.rotation += (1 * r2) } } stage = MainStage.go(/*{ loop: true }*/)