d3fc-random-data

Components for generating random data series based on stochastic processes.

Main D3FC package

Installing

1npm install @d3fc/d3fc-random-data

API Reference

Financial

The random financial data generator component generates open-high-low-close-volume financial data. Prices are calculated using the geometric Brownian motion generator.

1
2import { randomFinancial, randomSkipWeekends } from 'd3fc-random-data';
3
4const generator = randomFinancial()
5    .startDate(new Date(2016, 0, 1))
6    .startPrice(100)
7    .filter(randomSkipWeekends);
8
9generator(4);
10
11// [
12//   {
13//     date: 2016-01-01T00:00:00.000Z,
14//     open: 100,
15//     high: 100.37497903455065,
16//     low: 99.9344064016257,
17//     close: 100.13532170178823,
18//     volume: 974
19//   },
20//   {
21//     date: 2016-01-04T00:00:00.000Z,
22//     open: 100.2078374019404,
23//     high: 100.55251268471399,
24//     low: 99.7272105851512,
25//     close: 99.7272105851512,
26//     volume: 992
27//   },
28//   {
29//     date: 2016-01-05T00:00:00.000Z,
30//     open: 99.7272105851512,
31//     high: 101.06403178230532,
32//     low: 99.7272105851512,
33//     close: 101.00200313600685,
34//     volume: 835
35//   },
36//   {
37//     date: 2016-01-06T00:00:00.000Z,
38//     open: 101.00200313600685,
39//     high: 101.41129520567128,
40//     low: 100.50311227829566,
41//     close: 100.5536971451326,
42//     volume: 1021
43//   }
44// ]
45

# fc.randomFinancial()

Constructs a new financial data generator.

# randomFinancial(points)

Run the generator. Returns an array with points number of objects with date, open, high, low, close and volume properties.

# randomFinancial.startDate([value])

If value is specified, sets the start date to the specified Date object and returns this generator instance. If value is not specified, returns the current start date, which defaults to the value of new Date() when the generator was constructed.

# randomFinancial.startPrice([value])

If value is specified, sets the start price to the specified number and returns this generator instance. If value is not specified, returns the current start price, which defaults to 100.

# randomFinancial.interval([value])

If value is specified, sets the time increment to the specified d3 time interval and returns this generator instance. If value is not specified, returns the current interval, which defaults to d3_time.timeDay.

# randomFinancial.intervalStep([value])

If value is specified, sets the number of intervals that returned points should have dates offset by to the specified integer number and returns this generator instance. If value is not specified, returns the current number of intervals, which defaults to 1. Internally, this value is supplied to the step argument of an interval's offset function.

# randomFinancial.steps([value])

Get/Set the number of steps used by the geometric Brownian motion simulation per intervalStep number of intervals. A higher number gives a slower, but higher resolution simulation.

# randomFinancial.mu([value])

Get/Set the drift used by the geometric Brownian motion simulation.

# randomFinancial.sigma([value])

Get/Set the volatility used by the geometric Brownian motion simulation.

# randomFinancial.random([value])

Get/Set the random function used by the geometric Brownian motion simulation.

# randomFinancial.unitInterval([value])

If value is specified, sets the time interval used for units of mu and sigma to the specified d3 time interval and returns this generator instance. If value is not specified, returns the current interval, which defaults to d3_time.timeYear.

# randomFinancial.unitIntervalStep([value])

If value is specified, sets the integer number of intervals used for units of mu and sigma to the specified number and returns this generator instance. If value is not specified, returns the current interval, which defaults to 1. For example, to have trading year units of mu and sigma rather than calendar year, set unitIntervalStep to 252 and unitInterval to d3_time.timeDay.

# randomFinancial.volume([value])

If value is specified, sets the function used return a point's volume to the specified function and returns this generator instance. Can be specified as either a function mapping an output object to a number, or a number. If value is not specified, returns the current volume, which defaults to a function sampling integers from a normal distribution centred around 1000.

# randomFinancial.filter([value])

If value is specified, sets the filter function to the specified function and returns this generator instance. Only output objects d for which filter(d) returns true will be included in the output array. If value is not specified, returns the current filter function, which defaults to (d) => true. To skip weekends, supply the pre-defined filter fc_random_data.skipWeekends.

Financial Stream

Use the streaming interface to have successive calls to generate data keep track of the latest date and price.

1
2import { randomFinancial } from 'd3fc-random-data';
3
4const generator = randomFinancial()
5    .startDate(new Date(2016, 0, 1))
6    .startPrice(100);
7
8const stream = generator.stream();
9let data = [];
10
11data.push(stream.next());
12// data.length -> 1
13
14data = data.concat(stream.take(2));
15// data.length -> 3
16
17data = data.concat(stream.until(d => d.date > new Date(2016, 0, 10)));
18// data.length -> 10
19

# randomFinancial.stream()

Constructs a new stream from an existing financial data generator instance.

# stream.next()

Returns a single output object with date incremented from the latest returned output object's date according to the generator instance's interval and intervalStep properties, or with startDate if this is the first call.

# stream.take(number)

Returns an array of length number of output objects, each object with date incremented according to the generator instance's interval and intervalStep properties, starting with the latest returned output objects's incremented date, or with startDate if this is the first call.

# stream.until(comparison)

Returns the array of objects constructed by repeatedly generating a single output object with date incremented according to the generator instance's interval and intervalStep properties until a generated object satisfies the condition of the supplied comparison function, appending to the output array only if the condition is not satisfied.

Geometric Brownian Motion

The geometric Brownian motion component creates a series of values based on the Geometric Brownian Motion stochastic process.

1
2import { randomGeometricBrownianMotion } from 'd3fc-random-data';
3
4const generator = randomGeometricBrownianMotion()
5    .steps(10);
6
7generator(10);
8
9// [
10//   10,
11//   10.272847363463436,
12//   10.423881104466574,
13//   10.629316182766384,
14//   10.7209321393133,
15//   10.773722182206432,
16//   10.229636144307582,
17//   10.225282323984114,
18//   10.488138829847468,
19//   10.428118194568341,
20//   10.848822656937935
21// ]
22

# fc.randomGeometricBrownianMotion()

Constructs a new geometric Brownian motion generator.

# randomGeometricBrownianMotion(start)

Returns an array of price values following a geometric Brownian motion with the set drift and volatility, given a starting price of start. The first array value is the supplied start price, followed by steps number of values corresponding to the simulated price value at the end of each step.

# randomGeometricBrownianMotion.mu([value])

If value is specified, sets the percentage drift per period to the specified number and returns this generator instance. If value is not specified, returns the current drift, which defaults to 0.1.

# randomGeometricBrownianMotion.sigma([value])

If value is specified, sets the percentage volatility per period to the specified number and returns this generator instance. If value is not specified, returns the current volatility, which defaults to 0.1.

# randomGeometricBrownianMotion.period([value])

If value is specified, sets the interval length to the specified number of periods and returns this generator instance. If value is not specified, returns the current interval length, which defaults to 1.

randomGeometricBrownianMotion.steps([value])

If value is specified, sets the number of discrete steps to divide the interval into to the specified number and returns this generator instance. If value is not specified, returns the current number of steps, which defaults to 20.

randomGeometricBrownianMotion.random([value])

If value is specified, sets the function used for generating random numbers with a normal (Gaussian) distribution to the specified function and returns this generator instance. If value is not specified, returns the current random function, which defaults to d3_random.randomNormal.