Question

I need to generate a random number from the following data:

0-10 : 23%

10-80 : 50%

80-100 : 27%

How do I generate a random number from such information?

One way would be to fit a distribution but I have to do this for about hundred variables and I don’t want to fit 100 distributions. Any hints?

import random
a=[23, 73, 100]
b=[10, 80, 100]
rndval=awesomefunction(a,b)

Now, regarding awesomefunction(), I have absolutely no clue.

But, from what little I know, (and a very sloppy implementation)

temp_rand=random.uniform(0,100)
if(temp_rand<=23):
 rndval=random.uniform(0,10)
if(temp_rand<=73 && temp_rand>23):
 rndval=random.uniform(10,80)
if(temp_rand>73):
 rndval=random.uniform(80,100)

But IMHO, this is sloppy beyond measure.

Answer 1

You can have 2 random generators to achieve this. The first should generate a random double from 0 to 1.

For this generator, you should check to see if the random number is from 0 to .22, .23 to .73, or .74 to 1. If the number falls within the first range, you just run another random number generator that generates a number from 0 to 10, and that’s your number. The same goes for the second and third ranges.

Some high level pseudocode:

double firstRandomNumber = generateRandomNumberFromZeroToOne();
if (firstRandomNumber <= .22) {
    //generate random number from 0 to 10 and that's your number.
} else if (firstRandomNumber <= .73) {
    //generate random number from 10 to 80 and that's your number.
} else {
    //generate random number from 80 to 100 and that's your number.
}

A side note: Most languages already have random number generators that will generate a random number from 0 to 1 or an integer up to a specific range (0 to 10).