Part C$)$ Let's try to visualize how well our estimator actually fits the data.

To create a histogram of the "prints" data, one only needs to type hist$($prints$).$ In order to store the results in a way the autograder can interpret, we need

to store this in a variable like this: my$\_$hist$\_$var$=$hist$($prints$)$

Recall that the y$-$axis of a histogram can be scaled in many different ways. In order to make it so that the total area of each bar is the proportion of

observations observed in the corresponding bin, we need to add "prob$=$T$"($probability equals true$)$ to the histogram command. This is called a "density

histogram".

Finally, to control the way the bins break on the x$-$axis, we need to give a sequence of break values. As an example unrelated to this exercise, we might type

my$\_$hist$\_$var $=$ hist $($mydata$,$ prob$=$T$,$ breaks$=$seq $(1,8,1))$ to put breaks at $1,2,$dots,$8.$

For this exercise, create a density histogram of the prints values and store it in variable p$2.$c$.$ hist. Center the bars on the integers $0$ through $17$ by

making breaks at $-0\mathrm{.}5,0\mathrm{.}5,1\mathrm{.}5,$dots,$17\mathrm{.}5.$ Add the curve of a geometric distribution, with the probability of what you estimated in Part B$,$ ontop of your

histogram. Think about how well this curve approximates the distribution.

$[]$: p$2.$c$.$hist $=$ NA

# your code here

$[]$: # Hidden Test Cell

Part D$)$ what is the true underlying probability of this distribution? Save your answer as p$2.$d and round your answer to two decimal places. If it's not possible

to answer this question, save your answer as the string "not possible" $.$