Statistikk og Simulering

# Økt 19+. Analyse av predator/prey

## Session 9

#### 17.1. Session 9

Let’s explore some other meaningful characteristics, such as

• Expected population size at a troff.
• Expected population size at a peak.
• Number of births within an interval.
• Average age of animals dying within an interval.

The interval (in 3 and 4) should obviously cover a full period (a peak and a troff on the curve), and ideally and integer number of such periods. An odd fraction of a period could skew the distribution.

In the following exercise, you may have to rely on manual inspection to gather observations for the analysis, meaning that your sample may be small (10–20 observations). That is a good case for Student’s t-distribution.

Oppgåve 17.4 Choose at least two variables to estimate in the scenario where both predator and prey survive. Assume that it is normally distributed, and use Student’s $t$-distribution to estimate it.

Hint 1 In order to assess the period length, you have to identify the peaks and the troffs on the curve. The following matlab function may be useful:

1.
http://blogs.mathworks.com/pick/2008/05/09/finding-_local-_extrema/

Other variables one might want to estimate are the period length (time between subsequent peaks), probability of a species becoming extint, probability of a rabbit being eaten as opposed to dying of other causes, etc.

Oppgåve 17.5 Discuss in your group a number of numerical variables you might want to estimate, and the challenges involved in doing it.