# Diagonalizing the decision matrix

I've left you with some choices:

:::{table}
:align: center

| Command-line      | Notebook          | Either     |
| ----------------- | ----------------- | ---------- |
| ROOT C++          | ROOT C++          |            |
| Python with ROOT  | Python with ROOT  |            |
|                   |                   | RDataFrame |
:::

This tutorial is already too long, and I've taken longer than I
should have to offer you too many options. For simplicity, I've chosen
to present ROOT C++ on the command line in {ref}`cpath`, and Python with
ROOT in a Jupyter notebook in {ref}`pythonpath`. For {ref}`rdataframe`,
these choices don't matter much. 

If you choose to pursue one of the "off-diagonal" choices, you won't
have much trouble no matter which path you choose. You were previously introduced to
{ref}`ROOT C++ in a notebook <ckernel>`. To
run Python with ROOT on the command line (including magic commands), the
following will set you up on a Nevis particle-physics system:

    > conda activate /usr/nevis/conda/root
    > ipython

{ref}`cpath` and {ref}`pythonpath` present the same commands,
exercises, and footnotes.[^f69] You might even be able to do all of
these paths; once you've learned C++, Python is pretty easy. Tack on
{ref}`rdataframe` for total ROOT mastery![^nope]

[^nope]: Uh, no. Remember, it takes a lifetime to learn ROOT. While it
    may take more than two days to go through all three instructional
    paths, it won't take your entire lifetime. There's much more to
    ROOT than what's in this tutorial!

:::{figure-md} matrix_transform-fig
:class: align-center

<img src="https://imgs.xkcd.com/comics/matrix_transform.png" alt="xkcd matrix_transform">

<https://xkcd.com/184/> by Randall Munroe
:::

[^f69]: The xkcd cartoons in each path are different from the others, to give you an
    incentive to skim them all.