Take the 2-minute tour ×
Mathematics Stack Exchange is a question and answer site for people studying math at any level and professionals in related fields. It's 100% free, no registration required.

I want to generate/simulate a sequence of partial sums.

$$(X)_n(w) = \sum\limits_{i=1}^n (Y_i(w)-Y_{i-1}(w)),\text{ for }1 \leq n \leq 100$$

Let $W$ be a random variable such that: $W \thicksim N(0,1)$.

I know that: $\forall i, (Y_i(w)-Y_{i-1}(w))$ has the same distribution as $W$.

However, I don't know anything about how the individual $Y_i$ random variables are distributed, other than to say that they are, together with W, defined on the same probability space.

I'm really confused by a number of aspects about this question. In order to generate the partial sums, is it sufficient to take n independent draws out of a standard normal and then calculate the partial sums of these as my sequence $(X)_n(w)$?

But then, from the definition of the partial sums above, it is clear that $w$ is fixed throughout the partial sum. What does that mean for me?

My intuitive response is to ignore the fixed $w$. While I can draw on my knowledge of the distribution of $Y_i - Y_{i-1}$ to draw randomly from that distribution, I don't actually know anything about what their underlying functions look like. And so, for different values of $i$, $(Y_i(w) - Y_{i-1}(w)$ may map the same $w$ to vastly different values.

share|improve this question
add comment

Your Answer

 
discard

By posting your answer, you agree to the privacy policy and terms of service.

Browse other questions tagged or ask your own question.