mt19937
Create an array containing pseudorandom numbers generated using a 32-bit Mersenne Twister pseudorandom number generator.
Usage
var mt19937 = require( '@stdlib/random/array/mt19937' );
mt19937( len[, options] )
Returns an array containing pseudorandom integers on the interval [0, 4294967295]
.
var out = mt19937( 10 );
// returns <Float64Array>
The function has the following parameters:
- len: output array length.
- options: function options.
The function accepts the following options
:
- dtype: output array data type. Must be a real-valued data type or "generic". Default:
'float64'
.
By default, the function returns a Float64Array
. To return an array having a different data type, set the dtype
option.
var opts = {
'dtype': 'generic'
};
var out = mt19937( 10, opts );
// returns [...]
mt19937.normalized( len[, options] )
Returns an array containing pseudorandom numbers on the interval [0, 1)
with 53-bit precision.
var out = mt19937.normalized( 10 );
// returns <Float64Array>
The function has the following parameters:
- len: output array length.
- options: function options.
The function accepts the following options
:
- dtype: output array data type. Must be a real-valued floating-point data type or "generic". Default:
'float64'
.
By default, the function returns a Float64Array
. To return an array having a different data type, set the dtype
option.
var opts = {
'dtype': 'generic'
};
var out = mt19937.normalized( 10, opts );
// returns [...]
mt19937.factory( [options] )
Returns a function for creating arrays containing pseudorandom numbers generated using a 32-bit Mersenne Twister pseudorandom number generator.
var random = mt19937.factory();
var out = random( 10 );
// returns <Float64Array>
var len = out.length;
// returns 10
out = random.normalized( 10 );
// returns <Float64Array>
len = out.length;
// returns 10
The function accepts the following options
:
- seed: pseudorandom number generator seed.
- state: a
Uint32Array
containing pseudorandom number generator state. If provided, the function ignores theseed
option. - copy:
boolean
indicating whether to copy a provided pseudorandom number generator state. Setting this option tofalse
allows sharing state between two or more pseudorandom number generators. Setting this option totrue
ensures that a returned generator has exclusive control over its internal state. Default:true
. - idtype: default output array data type when generating integers. Must be a real-valued data type or "generic". Default:
'float64'
. - ndtype: default output array data type when generating normalized numbers. Must be a real-valued floating-point data type or "generic". Default:
'float64'
.
To seed the underlying pseudorandom number generator, set the seed
option.
var opts = {
'seed': 12345
};
var random = mt19937.factory( opts );
var out = random( 10, opts );
// returns <Float64Array>
The returned function accepts the following options
:
- dtype: output array data type. Must be a real-valued data type or "generic". This overrides the default output array data type.
The returned function has a normalized
method which accepts the following options
:
- dtype: output array data type. Must be a real-valued floating-point data type or "generic". This overrides the default output array data type.
To override the default output array data type, set the dtype
option.
var random = mt19937.factory();
var out = random( 10 );
// returns <Float64Array>
var opts = {
'dtype': 'generic'
};
out = random( 10, opts );
// returns [...]
mt19937.PRNG
The underlying pseudorandom number generator.
var prng = mt19937.PRNG;
// returns <Function>
mt19937.seed
The value used to seed the underlying pseudorandom number generator.
var seed = mt19937.seed;
// returns <Uint32Array>
mt19937.seedLength
Length of underlying pseudorandom number generator seed.
var len = mt19937.seedLength;
// returns <number>
mt19937.state
Writable property for getting and setting the underlying pseudorandom number generator state.
var state = mt19937.state;
// returns <Uint32Array>
mt19937.stateLength
Length of underlying pseudorandom number generator state.
var len = mt19937.stateLength;
// returns <number>
mt19937.byteLength
Size (in bytes) of underlying pseudorandom number generator state.
var sz = mt19937.byteLength;
// returns <number>
Notes
- Mersenne Twister is not a cryptographically secure PRNG, as the PRNG is based on a linear recursion. Any pseudorandom number sequence generated by a linear recursion is insecure, due to the fact that one can predict future generated outputs by observing a sufficiently long subsequence of generated values.
- Compared to other PRNGs, Mersenne Twister has a large state size (
~2.5kB
). Because of the large state size, beware of increased memory consumption when using thefactory()
method to create many Mersenne Twister PRNGs. When appropriate (e.g., when external state mutation is not a concern), consider sharing PRNG state. - A seed array of length
1
is considered equivalent to an integer seed equal to the lone seed array element and vice versa. - If PRNG state is "shared" (meaning a state array was provided during function creation and not copied) and one sets the underlying generator state to a state array having a different length, the function returned by the
factory
method does not update the existing shared state and, instead, points to the newly provided state array. In order to synchronize the output of the underlying generator according to the new shared state array, the state array for each relevant creation function and/or PRNG must be explicitly set. - If PRNG state is "shared" and one sets the underlying generator state to a state array of the same length, the PRNG state is updated (along with the state of all other creation functions and/or PRNGs sharing the PRNG's state array).
Examples
var logEach = require( '@stdlib/console/log-each' );
var mt19937 = require( '@stdlib/random/array/mt19937' );
// Create a function for generating random arrays originating from the same state:
var random = mt19937.factory({
'state': mt19937.state,
'copy': true
});
// Generate 3 arrays:
var x1 = random.normalized( 5 );
var x2 = random.normalized( 5 );
var x3 = random.normalized( 5 );
// Print the contents:
logEach( '%f, %f, %f', x1, x2, x3 );
// Create another function for generating random arrays with the original state:
random = mt19937.factory({
'state': mt19937.state,
'copy': true
});
// Generate a single array which replicates the above pseudorandom number generation sequence:
var x4 = random.normalized( 15 );
// Print the contents:
logEach( '%f', x4 );