Difference between revisions of "Aufgaben:Exercise 2.7: C Programs "z1" and "z2""
From LNTwww
Line 26: | Line 26: | ||
<quiz display=simple> | <quiz display=simple> | ||
− | { | + | {It is valid $M=4$ and $\text{p_array} = \big[0.2, \ 0.3, \ 0.4, \ 0.1 \big]$. |
− | <br> | + | <br>What result does the function $z1$ return if the random function returns the value $x = 0.75$ ? |
|type="{}"} | |type="{}"} | ||
− | $z1 \ = \ $ | + | $z1 \ = \ $ { 2 } |
− | { | + | {Which of the following statements are true regarding $z1$ ? |
|type="[]"} | |type="[]"} | ||
− | - | + | - You could omit the assignment $\text{x = random()}$ in line 5 and compare directly with $\text{random()}$ in line 8. |
− | + | + | + If all passed probabilities are equal, there would be faster program realizations than $z1$. |
− | + | + | + The return value $\text{random() = 0.2}$ leads to the result $z1= 1$. |
− | { | + | {Which of the following statements are true regarding $z2? |
|type="[]"} | |type="[]"} | ||
− | + | + | + The program generates a <i>binomially distributed</i> random variable. |
− | - | + | - The program generates a <i>poisson distributed</i> random variable. |
− | + | + | + With $I = 4$ für $z2$ the values $0, \ 1, \ 2, \ 3, \ 4$ are possible. |
− | + | + | + The inclusion of the mathematical library "'''math.h'''" is required because in $z2$ the function "'''pow'''" (exponentiate) is used. |
− | { | + | {What value is in $\text{p_array[2]}$ when called with $I = 4$ and $p = 0.25$? |
|type="{}"} | |type="{}"} | ||
− | $\text{p_array[2]} \ = | + | $\text{p_array[2]} \ = \ $ { 0.211 3% } |
Line 54: | Line 54: | ||
</quiz> | </quiz> | ||
− | === | + | ===Solution=== |
{{ML-Kopf}} | {{ML-Kopf}} | ||
− | '''(1)''' | + | '''(1)''' After the first iteration of the loop $(m = 0)$ the variable $\text{sum = 0.2}$, at the next iteration $(m = 1)$ holds $\text{sum = 0.2 + 0.3 = 0.5}$. |
− | *In | + | *In both cases, therefore, the variable $\text{sum} < x = 0.75$. |
− | * | + | *Only when $m = 2$ is the return '''(bin hier nicht sicher)''' condition satisfied: $0.9 > x$. Thus $\underline{z1 = 2}$. |
− | '''(2)''' | + | '''(2)''' The correct solutions are <u>solutions 2 and 3</u>: |
− | * | + | *Were we to dispense with the auxiliary variable $x$ and write in line 8 instead $\text{sum > random()}$ a new random value would be generated on each iteration of the loop, and $z1$ would then not have the desired properties. |
− | *$z1$ | + | *$z1$ works according to the diagram on the page "Generation of multistage random variables" in the theory section. There you find a much faster implementation for the case of equal probabilities $(1/M)$. |
− | * | + | *In the first run $(m = 0)$ in this case the return condition is not satisfied due to the less/equal–query; the output value is actually $z1 = 1$. |
− | '''(3)''' | + | '''(3)''' The correct <u>solutions are 1, 3, and 4</u>: |
− | * | + | *It results in a binomially distributed random variable, with the set of values $\{0, 1, 2, 3, 4\}$. |
− | * | + | *For the calculation of the probability ${\rm Pr}(z2 = 0) = (1 -p)^{I}$ one needs the mathematical library. |
− | * | + | *But exponentiation could also be realized by $I$–fold multiplication. |
− | '''(4)''' | + | '''(4)''' Because of line 6, the field element $\text{p_array[0]}$ before the program loop $(i = 0)$ contains the value $(1 -p)^{I}$. |
− | * | + | *In the first iteration $(i = 1)$ the following value is entered: |
− | :$$\text{p_array[1]}=\frac{ p\cdot I}{ 1- p}\cdot\text{p_array[0]}= | + | :$$\text{p_array[1]}=\frac{ p\cdot I}{ 1- p}\cdot\text{p_array[0]}= I\cdot p\cdot(1- p)^{ I- 1}={\rm Pr}(z2= 1) .$$ |
− | * | + | *In the second iteration $(i = 2)$ the probability for the result "$z2=2$" is calculated: |
− | :$$\text{p_array[2]}=\frac{p\cdot (I- 1)}{ 2\cdot ( 1- p)}\cdot\text{p_array[1]}= | + | :$$\text{p_array[2]}=\frac{p\cdot (I- 1)}{ 2\cdot ( 1- p)}\cdot\text{p_array[1]}= \left({ I \atop { 2}}\right)\cdot p^{\rm 2}\cdot( 1- p)^{\rm 2}={\rm Pr}( z2 = 2) .$$ |
− | * | + | *For $I= 4$ and $p = 0.25$ we get the following numerical value ⇒ "$4$ over $2$" $=6$: |
:$$\text{p_array[2]}={\rm Pr}( z 2=2)=6\cdot\frac{1}{16}\cdot\frac{9}{16} | :$$\text{p_array[2]}={\rm Pr}( z 2=2)=6\cdot\frac{1}{16}\cdot\frac{9}{16} | ||
\hspace{0.15cm}\underline{=0.211}.$$ | \hspace{0.15cm}\underline{=0.211}.$$ |
Revision as of 09:57, 21 December 2021
The two C programs given here are suitable for generating discrete random variables:
- The function $z1$ generates an $M$–stepped random variable with the value set $\{0, 1$, ... , $M-1\}$. The associated probabilities are passed in the array $\text{p_array}$ with property "Float" The function $\text{random()}$ returns equally distributed float–random variables between $0$ and $1$.
- A second function $z2$ (source code see below) returns a special probability distribution specified by the two parameters $I$ and $p$. This is done using the function $z1$.
Hints:
- The exercise belongs to the chapter Generation of Discrete Random Variables.
- In particular, reference is made to the page Generation of multistage random variables .
Questions
Solution
(1) After the first iteration of the loop $(m = 0)$ the variable $\text{sum = 0.2}$, at the next iteration $(m = 1)$ holds $\text{sum = 0.2 + 0.3 = 0.5}$.
- In both cases, therefore, the variable $\text{sum} < x = 0.75$.
- Only when $m = 2$ is the return (bin hier nicht sicher) condition satisfied: $0.9 > x$. Thus $\underline{z1 = 2}$.
(2) The correct solutions are solutions 2 and 3:
- Were we to dispense with the auxiliary variable $x$ and write in line 8 instead $\text{sum > random()}$ a new random value would be generated on each iteration of the loop, and $z1$ would then not have the desired properties.
- $z1$ works according to the diagram on the page "Generation of multistage random variables" in the theory section. There you find a much faster implementation for the case of equal probabilities $(1/M)$.
- In the first run $(m = 0)$ in this case the return condition is not satisfied due to the less/equal–query; the output value is actually $z1 = 1$.
(3) The correct solutions are 1, 3, and 4:
- It results in a binomially distributed random variable, with the set of values $\{0, 1, 2, 3, 4\}$.
- For the calculation of the probability ${\rm Pr}(z2 = 0) = (1 -p)^{I}$ one needs the mathematical library.
- But exponentiation could also be realized by $I$–fold multiplication.
(4) Because of line 6, the field element $\text{p_array[0]}$ before the program loop $(i = 0)$ contains the value $(1 -p)^{I}$.
- In the first iteration $(i = 1)$ the following value is entered:
- $$\text{p_array[1]}=\frac{ p\cdot I}{ 1- p}\cdot\text{p_array[0]}= I\cdot p\cdot(1- p)^{ I- 1}={\rm Pr}(z2= 1) .$$
- In the second iteration $(i = 2)$ the probability for the result "$z2=2$" is calculated:
- $$\text{p_array[2]}=\frac{p\cdot (I- 1)}{ 2\cdot ( 1- p)}\cdot\text{p_array[1]}= \left({ I \atop { 2}}\right)\cdot p^{\rm 2}\cdot( 1- p)^{\rm 2}={\rm Pr}( z2 = 2) .$$
- For $I= 4$ and $p = 0.25$ we get the following numerical value ⇒ "$4$ over $2$" $=6$:
- $$\text{p_array[2]}={\rm Pr}( z 2=2)=6\cdot\frac{1}{16}\cdot\frac{9}{16} \hspace{0.15cm}\underline{=0.211}.$$