To me it's an means to an end. I don't care if my solution takes 100ms instead of 1ms, it's the superior choice for me if it takes me 1 minute to do it instead of 10 minutes to learn something new.

True, but sometimes these 10 minutes help you to discover something new that will improve your code.

I had a few of these cases in my life:

- discovering optimized patterns in Perl, which led to code I could not understand the next day

- discovering decorators in Python, which led to better code

- discovering comprehensions in Python (a magical thing) that led to better code, except when I wanted to be too clever and ended up with Perl-like code

The flaw of human nature on display. And I don't mean that personal, not to you anyway, but to the human species.

Why? It seems very rational. Especially if you're just going to run it once to get a value and not as a part of some system.

Exactly, that's not a flaw, that's rational behaviour. Why design an intricate solution for a one-off. Using 15 times the amount of time it would have taken you manually to automate a pretty standard task is just stupid, though we all do it.

Do the first thing that works, don't overthink it.

I think the flaw is you misinterpreting the argument.

In my opinion, the bottom line with obvious caveats is this - Human-time is more valuable than CPU-time.

If you are shipping at scale then the calculus is different - Don't waste end-users' human-time and their cpu-time and/or server's cpu-time.

If you're writing code with a team the calculus is different - Use/Learn techniques and tools to reduce the teams' human-time wastage plus all the above.

If you're writing code just for yourself the calculus is different - Save your own human-time.

One doesn’t learn rock climbing to step over a brick, man.

Because when it's not a one off?

If that was a fair comment then we’d be writing all our code in assembly still.

I iterate over rows in pandas fairly often for plotting purposes. Anytime I want to draw something more complicated than a single point for each row, I find it's simple and straight-forward to just iterrows() and call the appropriate matplotlib functions for each. It does mean some plots that are conceptually pretty simple end up taking ~5 seconds to draw, but I don't mind. Is there really a better alternative that isn't super complicated? Keep in mind that I frequently change my mind about what I'm plotting, so simple code is really good (it's usually easier to modify) even if it's a little slower.

>That’s sad, as when you find yourself iterating over rows in pandas you’re almost invariably doing some wrong or very very sub optimally.

Humans writing code is suboptimal. I can't wait for the day when robots/AI do it for us. I just hope it leads to a utopia and not a dystopia.

I'm glad that DataFrames don't iterate by default. It's good design to make suboptimal features hard to access.