Astochastic model for a car insurance company's total cost of damages from traffic model for the costs of passenger car traffic accidents goes back to the work by van der lann and louter, a statistical accidents', journal of the royal statistical socioty (1986). for every sk-1,2.3dotss we denote by the random variable $ ks the us dollar amount of a damage from a policy holder's traffic accident which will occur during the year 2019. we assume that sx 1s, sx 2s.. is claim size of $1,000 usd. the (random) total number of accidents sns in 2019 is expected to be poisson distributed with 20 claims on average. an ii. d. independent and identically distributed) sequence of exponentially distributed random variables with an average t is assumed that the number of accidents is independent of the us doillar amount of damages for each accident. that is, the random variable sn$ is independent of the random variables s1, sx 2s the total costs for the insurance company by the end of 2019 will thus be given by the random sum note that the total number sns of accidents is random! the goal of the current exercise is to approximate the expected total costs $$imathbbleys. njs$s for the insuranpe company in 2019, and . the probabilities that the total cost will not exceed $ks usd, le via simulation. as usual, we start with loading some packages: tn [ ]: import numpy as np import math step 1: first, write a function which simulates the random variable ss n$. the output shoud just be a single scalar! hint use proper build-in functions from the numpy-package in your code in order to sample from a poisson distribution and from an exponential distribution: n 1: def randomsum(averageclaimsize, averagenumberof claims ## write your own code here return samplerandomsum