(practical-function)=
# Exercise 10: A more practical function

Your task is to repeat {ref}`more-physics` by writing a function to
compute $E_{loss}$:

:::{math}
E_{meas}^{2} = p_{x}^{2} + p_{y}^{2} + p_{z}^{2}

E_{loss} = E_{beam} - E_{meas}
:::

You'll create a new **`eloss`** column in the n-tuple, but that's the
*only* new column you're going to create. You are to write a function that
will take the needed columns from the n-tuple and return a value for
$E_{loss}$, use that function in a `Define`, then make a scatterplot
of this new column against **`zv`**.[^smart-aleck]

[^smart-aleck]: You can try to be clever and do this:

        eloss_df = dataframe.Define("eloss","ebeam-sqrt(px*px+py*py+pz*pz)")

    But you know that's not what I'm asking for. For a real physics
    project, you're going to be asked to performed calculations that
    can't be shoved into a simple one-line text string. Write an
    actual function so you can learn how it's done.

    No, you don't have to use a lambda expression. Unless, of course, you want to
    show how cool you are.

:::{hint}
Look at the available columns in the n-tuple. Compare that with above formulas. Which of those columns
will you need to compute those equations? How many arguments will your function take?
:::

:::{figure-md} equations-fig
:align: center

<img src="https://imgs.xkcd.com/comics/equations.png" alt="xkcd equations" width="75%">

<https://xkcd.com/2034/> by Randall Munroe
:::