Functions on an Outcome Space

Interact

Random sampling can be thought of as repeated random trials, and therefore many outcome spaces consist of sequences. An outcome space representing two tosses of a coin is

If you were tossing 10 times, the outcome space would consist of the $2^{10}$ sequences of 10 elements where each element is H or T. The outcomes are a pain to list by hand, but computers are good at saving us that kind of pain.

Product Spaces

The product of two sets $A$ and $B$ is the set of all pairs $(a, b)$ where $a \in A$ and $b \in B$. This concept is exactly what we need to describe spaces representing multiple trials.

For example, the space representing the outcome of one toss of a coin is $ \Omega_1 = { \text{H, T}}$. The product of $\Omega_1$ with itself is the set of pairs (H, H), (H, T), (T, H) and (T, T), which you will recognize as the outcomes of two tosses of a coin. The product of this new space and $\Omega_1$ is the space representing three tosses. And so on.

The Python module itertools contains a function product that constructs product spaces. Let’s import it.

from itertools import product

To see how product works, we will start with the outcomes of one toss of a coin. We are creating an array using make_array but you could use any other way of creating an array or list.

one_toss = make_array('H', 'T')

To use product, we have to specify the base space and the number of repetitions, and then covert the result to a list.

two_tosses = list(product(one_toss, repeat=2))
two_tosses
[('H', 'H'), ('H', 'T'), ('T', 'H'), ('T', 'T')]

For three tosses, just change the number of repetitions:

three_tosses = list(product(one_toss, repeat=3))
three_tosses
[('H', 'H', 'H'),
 ('H', 'H', 'T'),
 ('H', 'T', 'H'),
 ('H', 'T', 'T'),
 ('T', 'H', 'H'),
 ('T', 'H', 'T'),
 ('T', 'T', 'H'),
 ('T', 'T', 'T')]

A probability space is an outcome space accompanied by the probabilities of all the outcomes. If you assume all eight outcomes of three tosses are equally likely, the probabilities are all 1/8:

three_toss_probs = (1/8)*np.ones(8)

The corresponding probability space:

three_toss_space = Table().with_columns(
    'omega', three_tosses,
    'P(omega)', three_toss_probs
)
three_toss_space
omega P(omega)
['H' 'H' 'H'] 0.125
['H' 'H' 'T'] 0.125
['H' 'T' 'H'] 0.125
['H' 'T' 'T'] 0.125
['T' 'H' 'H'] 0.125
['T' 'H' 'T'] 0.125
['T' 'T' 'H'] 0.125
['T' 'T' 'T'] 0.125

Product spaces get large very quickly. If you roll a die 5 times, there are almost 8,000 possible outcomes:

6**5
7776

But we have product so we can still list them all! Here is a probability space representing 5 rolls of a die.

die = np.arange(1, 7, 1)

five_rolls = list(product(die, repeat=5))  # All possible results of 5 rolls

five_rolls_probs = (1/6**5)**np.ones(6**5)  # Each result has chance 1/6**5

five_rolls_space = Table().with_columns(
   'omega', five_rolls,
    'P(omega)', five_rolls_probs
)

five_rolls_space
omega P(omega)
[1 1 1 1 1] 0.000128601
[1 1 1 1 2] 0.000128601
[1 1 1 1 3] 0.000128601
[1 1 1 1 4] 0.000128601
[1 1 1 1 5] 0.000128601
[1 1 1 1 6] 0.000128601
[1 1 1 2 1] 0.000128601
[1 1 1 2 2] 0.000128601
[1 1 1 2 3] 0.000128601
[1 1 1 2 4] 0.000128601

... (7766 rows omitted)

A Function on the Outcome Space

Suppose you roll a die five times and add up the number of spots you see. If that seems artificial, be patient for a moment and you’ll soon see why it’s interesting.

The sum of the rolls is a numerical function on the outcome space $\Omega$ of five rolls. The sum is thus a random variable. Let’s call it $S$. Then, formally, The range of $S$ is the integers 5 through 30, because each die shows at least one spot and at most six spots. We can also use the equivalent notation

From a computational perspective, the elements of $\Omega$ are in the column omega of five_roll_space. Let’s apply this function and create a larger table.

five_rolls_sum = five_rolls_space.with_column(
    'S(omega)', five_rolls_space.apply(sum, 'omega')
).move_to_end('P(omega)')

five_rolls_sum
omega S(omega) P(omega)
[1 1 1 1 1] 5 0.000128601
[1 1 1 1 2] 6 0.000128601
[1 1 1 1 3] 7 0.000128601
[1 1 1 1 4] 8 0.000128601
[1 1 1 1 5] 9 0.000128601
[1 1 1 1 6] 10 0.000128601
[1 1 1 2 1] 6 0.000128601
[1 1 1 2 2] 7 0.000128601
[1 1 1 2 3] 8 0.000128601
[1 1 1 2 4] 9 0.000128601

... (7766 rows omitted)

We now have every possible outcome of five rolls of a die, along with its total number of spots. You can see that the first row of the table shows the smallest possible number of spots, corresponding to all the rolls showing 1 spot. The 7776th row shows the largest:

five_rolls_sum.take(7775)
omega S(omega) P(omega)
[6 6 6 6 6] 30 0.000128601

All the other values of $S$ are between these two extremes.

Functions of Random Variables

A random variable is a numerical function on $\Omega$. Therefore by composition, a numerical function of a random variable is also a random variable.

For example, $S^2$ is a random variable, calculated as follows:

Thus for example $S^2(\text{[6 6 6 6 6]}) = 30^2 = 900$.

Events Determined by $S$

From the table five_rolls_sum it is hard to tell how many rows show a sum of 6, or 10, or any other value. To better understand the properties of $S$, we have to organize the information in five_rolls_sum.

For any subset $A$ of the range of $S$, define the event ${S \in A}$ as

That is, ${ S \in A}$ is the collection of all $\omega$ for which $S(\omega)$ is in $A$. In terms of the table, the set consists of the values of $\omega$ in all the rows in which the sum is in $A$.

Try out the definition in a special case. Take $A = {5, 30}$. Then ${S \in A}$ if and only if either all the rolls show 1 spot or all the rolls show 6 spots. So

It is natural to ask about the chance the sum is a particular value, say 10. That’s not easy to read off the table, but we can access the corresponding rows:

five_rolls_sum.where('S(omega)', are.equal_to(10))
omega S(omega) P(omega)
[1 1 1 1 6] 10 0.000128601
[1 1 1 2 5] 10 0.000128601
[1 1 1 3 4] 10 0.000128601
[1 1 1 4 3] 10 0.000128601
[1 1 1 5 2] 10 0.000128601
[1 1 1 6 1] 10 0.000128601
[1 1 2 1 5] 10 0.000128601
[1 1 2 2 4] 10 0.000128601
[1 1 2 3 3] 10 0.000128601
[1 1 2 4 2] 10 0.000128601

... (116 rows omitted)

There are 126 values of $\omega$ for which $S(\omega) = 10$. Since all the $\omega$ are equally likely, the chance that $S$ has the value 10 is 126/7776.

We will usually be informal with notation and write ${ S = 10 }$ instead of ${ S \in {10} }$: