Class MTRandom

Uniform random number generator object.

Instantiate a uniform random number generator. This should be the Mersenne Twister based generator that is standard in Python (at least up to version 2.6).

Generates floating point numbers in the range [0-1].

If neither seed nor state are supplied, this will use the current time to generate a seed.

NB To set the matlab random number generator to the same known state as the python generator, you need to convert the signed values of the seed into unsigned values as follows:

>> seed = [get me from file etc..]; >> seed(seed < 0 ) = seed(seed < 0) + (2^32); >> rand('twister', seed);

Parameters:

• seed - (int) single integer seed value
• state - (int array) length 624 SIGNED 2-byte integer vector that completely captures the state of the random number generator. Note that these are SIGNED values -- matlab expects UNSIGNED values; see notes below. This comes from .getstate() method below

Overrides: object.__init__

Returns:

(array) state vector

Parameters:

• state - (array) state vector from getstate()

Returns:

nothing

Generate random numbers.

With no arguments generates a single floating point random number on the range [0-1]. If count is specified, then returns a list of count random numbers.

Parameters:

• count - (None or int) if not None, number of numbers to generate

Returns:

(float) random floats on range [0-1]