RND
Returns a random double precision number in the range [0, 1)
Syntax:
Usage:
result = Rnd( seed )
Parameters:
seed
Optional single argument. If seed has a value of zero (0.0), the last random number generated is repeate. For any other number a new random number is returned. With the QB-compatible algorithm, a negative number fully reseeds the generator. The default for no argument is to return a new random number.
Return Value:
Returns the random number generated.
Description:
Returns a number of type double in the range [0, 1) (i.e. 0 <= Rnd < 1), based on a random seed (see Randomize).
Rnd can use a variety of different algorithms - see Randomize for details of the default and selectable algorithms.
Rnd will return the same sequence of numbers every time a program is run. This sequence can be changed by reseeding the generator.
Note:
Rnd can use a variety of different algorithms - see Randomize for details of the default and selectable algorithms.
Rnd will return the same sequence of numbers every time a program is run. This sequence can be changed by reseeding the generator.
Note:
Rnd is thread-safe (by using an internal mutex), but not thread specific. Used in the context of a multi-threaded program, only the execution speed is significantly slowed down compared to a single thread call context.
Unfinished structures for other random number generators are also available in the standard header "fbprng.bi".
Unfinished structures for other random number generators are also available in the standard header "fbprng.bi".
Examples:
'' Function to a random number in the range [first, last), or {first <= x < last}.
Function rnd_range (first As Double, last As Double) As Double
Function = Rnd * (last - first) + first
End Function
'' seed the random number generator, so the sequence is not the same each time
Randomize
'' prints a random number in the range [0, 1), or {0 <= x < 1}.
Print Rnd
'' prints a random number in the range [0, 10), or {0 <= x < 10}.
Print Rnd * 10
'' prints a random integral number in the range [1, 10], or {1 <= n <= 10}.
'' (because: 0 <= Rnd * 10 < 10)
Print Int(Rnd * 10) + 1
'' prints a random integral number in the range [69, 420], or {69 <= n <= 420}.
'' (because: 69 <= rnd_range(69, 421) < 421)
Print Int(rnd_range(69, 421))
Function rnd_range (first As Double, last As Double) As Double
Function = Rnd * (last - first) + first
End Function
'' seed the random number generator, so the sequence is not the same each time
Randomize
'' prints a random number in the range [0, 1), or {0 <= x < 1}.
Print Rnd
'' prints a random number in the range [0, 10), or {0 <= x < 10}.
Print Rnd * 10
'' prints a random integral number in the range [1, 10], or {1 <= n <= 10}.
'' (because: 0 <= Rnd * 10 < 10)
Print Int(Rnd * 10) + 1
'' prints a random integral number in the range [69, 420], or {69 <= n <= 420}.
'' (because: 69 <= rnd_range(69, 421) < 421)
Print Int(rnd_range(69, 421))
- Before fbc 1.10.0 (for fbc 1.08.0 and fbc 1.09.0), "fbprng.bi" was named "fbmath.bi".
- Before fbc 1.08.0:
Rnd was not thread-safe (many values temporarily returned in duplicate for a same thread).
The standard "fbmath.bi" header for available algorithms did not exist.
The standard "fbmath.bi" header for available algorithms did not exist.
Dialect Differences:
The default algorithm used depends on the current dialect in use:
- With the -lang fb dialect, a 32 bit Mersenne Twister function with a granularity of 32 bits is used.
- With the -lang qb dialect, a function giving the same output as Rnd in QB is used. The granularity is 24 bits.
- With the -lang deprecated and -lang fblite dialects, the function in the C runtime available in the system is used. The function available in Win32 has a granularity of 15 bits, and 32 bits in Linux and DOS.
Differences from QB:
- None, if compiled in the -lang qb dialect. Other dialects can also use the same seeding and generating algorithms by calling Randomize with the appropriate parameter.
- For the non-QB-compatible algorithms, if the optional argument is less than 0, it has the same meaning as passing an argument of 1.
See also:
Back to Math
