CODEWITHSUNDEEP
c++stlrandomc++11srand
Joey Arnold Andres
Joey Arnold Andres

Reputation: 368

Rand_Max*(max-min)+min << what is that?

generate(vec.begin(), vec.end(), [=](){return static_cast<T>(static_cast<double>(ran())
/RAND_MAX*(max-min)+min); });

Problem: RAND_MAX*(max-min)+min);

Ok, so I know the algorithms, lambda expressions, and, capture clause in this thing. My question is quite ridiculous to all of that. What is the bolded text above mean. I mean, I know its part of the random value generation process. But don't know exactly what the hell is going on. So can someone pls break down that tinny tiny little code.

Upvotes: 2

Views: 5373

Answers (5)

Tiago Peczenyj
Tiago Peczenyj

Reputation: 4623

the ran() function returns a random value between min and max ?

but sometimes we need a random value between A and B (min and max). So we can adjust the result for it.

a is double, so we use a static_cast! 
a = rand()      ; 0     <=   a   <= RAND_MAX
a = a/RAND_MAX  ; 0     <=   a   <= 1
a = B * a       ; 0     <=   a   <= B
a = min +a      ; 0+min <=   a   <= B+min

to produce min <= a <= max, 

B+min = max
B = max-min

in other hands

a = rand()/RAND_MAX*(max-min) + min

is a random number between min and max.

Upvotes: 0

smocking
smocking

Reputation: 3709

static_cast<double>(ran())/RAND_MAX*(max-min)+min);

I'm assuming you mistyped rand(), which returns a pseudorandom integer from 0 to RAND_MAX. Let's rewrite that in a way that clarifies the precedence a bit:

(T) ( (((double) rand() / RAND_MAX) * (max-min) ) + min

So what it does is:

  1. rand(): take a random integer between 0 and RAND_MAX
  2. (double) / RAND_MAX: divide as double by RAND_MAX, yielding a uniformly distributed double between 0 and 1:
  3. * (max-min): multiply by the range (max-min), yielding a double from 0 to (max-min)
  4. +min: add the minimum to yield a random double between min and max
  5. static_cast<T>: cast it back to the original type

The result is a uniformly distributed random number of type T between min and max.

Upvotes: 8

Eugen Constantin Dinca
Eugen Constantin Dinca

Reputation: 9140

You'll need to look at the entire expression: static_cast<double>(rand()) /RAND_MAX*(max-min)+min). Which with explicit grouping looks like: (static_cast<double>(rand()) /RAND_MAX)*(max-min)+min).
The first group returns a random value between 0 & 1 since rand() returns a value in the range 0 to RAND_MAX. The second group translates the 0 to 1 range to a min to max range.

Upvotes: 2

guga
guga

Reputation: 724

The expression static_cast<double>rand()/RAND_MAX creates a number between 0.0 and 1.0
When you multiply it by (max-min), you change the range, and when you add to min you shift the range. So, after that expression you have a random number (double) that ranges from min to max.

Upvotes: 3

Robot Woods
Robot Woods

Reputation: 5687

It's a random function limited on the downside by min (because the rand piece could return zero) and limited to max because even if it returned 100% of max-min, and added to min, you'd be at max

Upvotes: 2

Related Questions