Probability related to cumulative distribution function

4 replies · 2K views · last reply
songoku
Messages
2,522
Reaction score
395
Homework Statement
Please see below
Relevant Equations
F(X) = P(X ≤ x)
Typeset problem statement defining a random variable X with piecewise CDF F_X(x): 0 for x < 0, i^2/N^2 for x in [i, i+1), 1 for x >= N, followed by parts (i) and (ii).


I have tried to answer all the questions but I am not that sure with my answer.

Hand-drawn step graph of a cumulative distribution function F_X(x) against x, with a step of height i^2/N^2 from x = i to x = i+1 and a step up to 1 at x = N.

That's the graph of ##F_X (x)## (I think)

(i) P (X ≤ i) = ##\frac{i^2}{N^2}## and P(X < i) = 0
All of these are based on the graph

(ii) P(X = i) = P(X ≤ i) - P(X < i) = ##\frac{i^2}{N^2}##

Are my answers correct? Thanks
 
songoku said:
Homework Statement: Please see below
Relevant Equations: F(X) = P(X ≤ x)

View attachment 333620

I have tried to answer all the questions but I am not that sure with my answer.

View attachment 333621
That's the graph of ##F_X (x)## (I think)

(i) P (X ≤ i) = ##\frac{i^2}{N^2}## and P(X < i) = 0
All of these are based on the graph
If there are several integers between 0 and i, they have positive probability values. so P(X<i) > 0.
songoku said:
(ii) P(X = i) = P(X ≤ i) - P(X < i) = ##\frac{i^2}{N^2}##

Are my answers correct? Thanks
No. Notice that your diagram only has one i < N, but there might be several others. Also, the sum of the probabilities must equal 1, so your diagram is missing a lot of probability.
 
FactChecker said:
If there are several integers between 0 and i, they have positive probability values. so P(X<i) > 0.

No. Notice that your diagram only has one i < N, but there might be several others. Also, the sum of the probabilities must equal 1, so your diagram is missing a lot of probability.
Ah, I see. Now I understand the question

Revised attempt:
(i)
$$P (X ≤ i) = \frac{i^2}{N^2}$$

$$P(X < i) =
\begin{cases}
0 & \text{if } i= 0 \\
\frac{(i-1)^2}{N^2} & \text{if } i>0
\end{cases}
$$

(ii)
$$P(X = i) =
\begin{cases}
0 & \text{if } i= 0 \\
\frac{2i-1}{N^2} & \text{if } i>0
\end{cases}
$$

For P(X < i) and P(X = i), is there an answer not involving piecewise function? Thanks
 
just resuming your effort : using Heaviside step function H
[tex]F_X(x)=\frac{1}{N^2}\sum_{i=1}^N (i^2-(i-1)^2)H_1(x-i)[/tex]
Probability density is by differentiation
[tex]p(x)=\frac{1}{N^2}\sum_{i=1}^N (i^2-(i-1)^2)\delta(x-i)[/tex]
Probability for digits are by integrating p(x) around x=i
[tex]P(i)=\frac{i^2-(i-1)^2}{N^2}=\frac{2i-1}{N^2}[/tex]
for ##1 \leq i \leq N##. Otherwise p(x)=0, P(i)=0.
 
Last edited:
Thank you very much for the help and explanation FactChecker and anuttarasammyak