function-plot

A 2d function plotter powered by d3

Function Plot is a powerful library built on top of D3.js whose purpose is to render functions with little configuration (think of it as a little clone of Google's plotting utility: y = x * x

The library currently supports interactive line charts and scatterplots, whenever the graph scale is modified the function is evaluated again with the new bounds, result: infinite graphs!

homepage

Install

$ npm install --save function-plot

Usage with browserify

var d3 = window.d3
var functionPlot = require('function-plot');
functionPlot({
  // options below
})

Example

All the available options are described in the homepage

API

var functionPlot = require('function-plot');

instance = functionPlot(options)

params, All the params are optional unless otherwise stated

• options {Object}
  • target {string} the selector of the parent element to render the graph to
  • title {string} If set the chart will have it as a title on the top
  • xAxis {Object}
    • type {string} default value: 'linear' the scale of this axis, possible values linear|log
    • domain {number[]} initial ends of the axis
    • label {string} label to show near the axis
  • yAxis {Object}
    • type {string} default value: 'linear' the scale of this axis, possible values linear|log
    • domain {number[]} initial ends of the axis
    • label {string} label to show near the axis
  • disableZoom {boolean} true to disable drag and zoom on the graph
  • grid {boolean} true to show a grid
  • tip {object} configuration passed to lib/tip, it's the helper shown on mouseover on the closest function to the current mose position
    • xLine {boolean} true to show a line parallel to the X axis on mouseover
    • yLine {boolean} true to show a line parallel to the Y axis on mouseover
    • renderer {function} Function to be called to define custom rendering on mouseover, called with the x and f(x) of the function which is closest to the mouse position (args: x, y)
  • annotations {Object[]} An array defining parallel lines to the y-axis or the x-axis
    • x {number} x-coordinate of the line parallel to the y-axis
    • y {number} y-coordinate of the line parallel to the x-axis
    • text {string} text shown next to the parallel line
  • data {array} required An array defining the functions to be rendered
  • plugins {array} An array describing plugins to be run when the graph is initialized, check out the examples on the main page

options.data {Array}

An array of objects, each object contains info of a function to render and can have the following options

• title {string} title of the function
• skipTip {boolean=false} true to make the tip ignore this function
• range {number[]=[-Infinity, Infinity]} an array with two numbers, the function will only be evaluated with values that belong to this interval
• nSamples {number} The number of values to be taken from range to evaluate the function, note that if interval-arithmetic is used the function will be evaluated with intervals instead of single values
• graphType {string='interval'} The type of graph to render, available values are interval|polyline|scatter
• fnType {string='linear'} The type of function to render, available values are linear|parametric|implicit|polar|points|vector
• sampler {string='interval'} The sampler to take samples from range, available values are interval|builtIn
  • NOTE: builtIn should only be used when graphType is polyline|scatter
  • NOTE: when math.js is included in the webpage it will be used instead of the bundled sampler

Additional style related options

• color {string} color of the function to render
• attr {Object} additional attributes set on the svg node that represents this datum
• closed {boolean=false} (only if graphType: 'polyline' or graphType: 'scatter') True to close the path, for any segment of the closed area graph y0 will be 0 and y1 will be f(x)

When derivative {Object} is present on a datum

• derivative.fn {string|Function} The derivative of fn
• derivative.x0 {number} The abscissa of the point which belongs to the curve represented by fn whose tangent will be computed (i.e. the tangent line to the point x0, fn(x0))
• derivative.updateOnMouseMove {boolean} True to compute the tangent line by evaluating derivative.fn with the current mouse position (i.e. let x0 be the abscissa of the mouse position transformed to local coordinates, the tangent line to the point x0, fn(x0))

When secants {Array} is present on a datum

• secants[i].x0 {number} The abscissa of the first point
• secants[i].x1 {number} (optional if updateOnMouseMove is set) The abscissa of the second point
• secants[i].updateOnMouseMove {boolean} (optional) True to update the secant line by evaluating fn with the current mouse position (x0 is the fixed point and x1 is computed dynamically based on the current mouse position)

if fnType: 'linear' (default)

• fn {string|Function} the function that represents the curve, this function is evaluated with values which are inside range

if fnType: 'parametric'

• x {string|Function} the x-coordinate of a point to be sampled with a parameter t
• y {string|Function} the y-coordinate of a point to be sampled with a parameter t
• range = [0, 2 * Math.PI] {Array} the range property in parametric equations is used to determine the possible values of t, remember that the number of samples is set in the property samples

if fnType: 'polar'

• r {string|Function} a polar equation in terms of theta
• range = [-Math.PI, Math.PI] the range property in polar equations is used to determine the possible values of theta, remember that the number of samples is set in the property samples

if fnType: 'implicit'

• fn {string|Function} a function which needs to be expressed in terms of x and y

NOTE: implicit functions can only be rendered using interval-arithmetic

if fnType: 'points'

• points {Array} an array of 2-number array which hold the coordinates of the points to render

NOTE: make sure your type of graph is either scatter or polyline

if fnType: 'vector'

• vector {Array} an 2-number array which has the ends of the vector
• offset {Array=[0, 0]} (optional) vector's offset

instance

• instance.id {string} a random generated id made out of letters and numbers
• instance.linkedGraphs {array} array of function-plot instances linked to the events of this instance, i.e. when the zoom event is dispatched on this instance it's also dispatched on all the instances of this array
• instance.meta {object}
  • instance.meta.margin {object} graph's left,right,top,bottom margins
  • instance.meta.width {number} width of the canvas (minus the margins)
  • instance.meta.height {number} height of the canvas (minus the margins)
  • instance.meta.xScale {d3.scale.linear} graph's x-scale
  • instance.meta.yScale {d3.scale.linear} graph's y-scale
  • instance.meta.xAxis {d3.svg.axis} graph's x-axis
  • instance.meta.yAxis {d3.svg.axis} graph's y-axis
• instance.root {d3.selection} svg element that holds the graph (canvas + title + axes)
• instance.canvas {d3.selection} g.canvas element that holds the area where the graphs are plotted (clipped with a mask)

Events

An instance can subscribe to any of the following events by doing instance.on([eventName], callback), events can be triggered by doing instance.emit([eventName][, params])

• mouseover fired whenever the mouse is over the canvas
• mousemove fired whenever the mouse is moved inside the canvas, callback params: a single object {x: number, y: number} (in canvas space coordinates)
• mouseout fired whenever the mouse is moved outside the canvas
• before:draw fired before drawing all the graphs
• after:draw fired after drawing all the graphs
• zoom:scaleUpdate fired whenever the scale of another graph is updated, callback params xScale, yScale (x-scale and y-scale of another graph whose scales were updated)
• tip:update fired whenever the tip position is updated, callback params x, y, index (in canvas space coordinates, index is the index of the graph where the tip is on top of)
• eval fired whenever the sampler evaluates a function, callback params data (an array of segment/points), index (the index of datum in the data array), isHelper (true if the data is created for a helper e.g. for the derivative/secant)

The following events are dispatched to all the linked graphs

• all:mouseover same as mouseover but it's dispatched in each linked graph
• all:mousemove same as mousemove but it's dispatched in each linked graph
• all:mouseout same as mouseout but it's dispatched in each linked graph
• all:zoom:scaleUpdate same as zoom:scaleUpdate but it's dispatched in each linked graph
• all:zoom fired whenever there's scaling/translation on the graph, dispatched on all the linked graphs

When the definite-integral plugin is included the instance will fire the following events

• definite-integral
  • datum {object} The datum whose definite integral was computed
  • i {number} The index of the datum in the data array
  • value {number} The value of the definite integral
  • a {number} the left endpoint of the interval
  • b {number} the right endpoint of the interval

Recipes

Evaluate a function at some value x

var y = functionPlot.eval.builtIn(datum, fnProperty, scope)

Where datum is an object that has a function to be evaluated in the property fnProperty , to eval this function we need an x value which is sent through the scope

e.g.

var datum = {
  fn: 'x^2'
}
var scope = {
  x: 2
}
var y = functionPlot.eval.builtIn(datum, 'fn', scope)

Every element of the data property sent to functionPlot is saved on instance.options.data, if you want to get the evaluated values of all the elements here run

var instance = functionPlot( ... )
instance.options.data.forEach(function (datum) {
  var datum = {
    fn: 'x^2'
  }
  var scope = {
    // a value for x
    x: 2
  }
  var y = functionPlot.eval.builtIn(datum, 'fn', scope)
}

Programmatic zoom

Just call instance.programmaticZoom with the desired x and y domains

var instance = functionPlot( ... )
var xDomain = [-3, 3]
var yDomain = [-1.897, 1.897]
instance.programmaticZoom(xDomain, yDomain)

Maintain aspect ratio

Given the xDomain values you can compute the corresponding yDomain values to main the aspect ratio between the axes

function computeYScale (width, height, xScale) {
  var xDiff = xScale[1] - xScale[0]
  var yDiff = height * xDiff / width
  return [-yDiff / 2, yDiff / 2]
}

var width = 800
var height = 400

// desired xDomain values
var xScale = [-10, 10]

functionPlot({
  width: width,
  height: height,
  xDomain: xScale,
  yDomain: computeYScale(width, height, xScale),

  target: '#demo',
  data: [{
    fn: 'x^2',
    derivative: {
      fn: '2x',
      updateOnMouseMove: true
    }
  }]
})

Changing the format of the values shown on the axes

var instance = functionPlot({
  target: '#complex-plane',
  xLabel: 'real',
  yLabel: 'imaginary'
})
// old format
var format = instance.meta.yAxis.tickFormat()
var imaginaryFormat = function (d) {
  // new format = old format + ' i' for imaginary
  return format(d) + ' i'
}
// update format
instance.meta.yAxis.tickFormat(imaginaryFormat)
// redraw the graph
instance.draw()

Styling

Selectors (sass)

.function-plot {
  .x.axis {
    .tick {
      line {
        // grid's vertical lines
      }
      text {
        // x axis labels
      }
    }
    path.domain {
      // d attribute defines the graph bounds
    }
  }

  .y.axis {
    .tick {
      line {
        // grid's horizontal lines
      }
      text {
        // y axis labels
      }
    }
    path.domain {
      // d attribute defines the graph bounds
    }
  }
}

Development

After cloning the repo and running npm install

node site.js    // generate the examples shown on index.html
npm start

Open 127.0.0.1:5555 and that's it! Local development server powered beefy

Development page: 127.0.0.1:5555/playground.html

Deploy steps:

• npm run dist (make sure to commit the dist files after this command)
• npm version major|minor|patch
• git push origin master
• npm run deploy
• npm publish

License

2015 MIT © Mauricio Poppe