var rnorm = function(bias, pow) {
    // using a mu just to avoid the special case of 0 centering
    return bias + (Math.random() > 0.5 ? 1 : -1) * Math.pow(Math.abs(
                Math.random() + Math.random() + Math.random()
                + Math.random() + Math.random() + Math.random() - 3) / 3, pow)
}

tserLabels = R.always(['Insulin-like growth factor', 'Von Willebrand Factor', 'Voltage-gated 6T & 1P',
    'Mechanosensitive ion ch.', 'GABAA receptor positive ', 'Epidermal growth factor',
    'Signal recognition particle'].slice(0,7))

var samples = flyd.stream()

function sampler(time) {
    return tserLabels().map(function(d, i) {return {key: time, value: rnorm(15, 0.25 + (2 - 0.25) * i / tserLabels().length)}})
}

var historyContextLength = 64
var historyShownLength = 16
var initialLength = 16

var initialHistory = R.map(sampler)(R.range(0, initialLength)) // ensuring 2 data points to give sufficient input to the Y scale

function generateSample(time) {
    samples(sampler(time))
}

var time = initialLength

var samplesHistoricalContext = flyd.stream([samples], function() {
    var newHistory = R.concat(samplesHistoricalContext(), [samples()])
    samplesHistoricalContext(R.slice(-historyContextLength, newHistory.length, newHistory))
})(initialHistory)

var pause = false

var timeCadence = 100

window.setInterval(function(){if(!pause) generateSample(time++)}, timeCadence)

function tserMaker(history) {
    var tserLength = history.length
    var range = R.range(0, tserLength)
    var tsers = tserLabels().map(function(d, i) {
        var full = R.map(function(time) {
            return history[time][i]
        })(range)
        return {
            key: d,
            contextValue: full,
            value: R.slice(-historyShownLength, full.length, full)
        }
    })
    return tsers
}

var model = flyd.stream([samplesHistoricalContext], function() {
    var history = samplesHistoricalContext()
    return tserMaker(history)
})