Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Lessons Learned from Two Years as a Data Scientist

dawndrain.github.io

Post reply on HN
Print view View original

51–60 of 89 posts

Re: Lessons Learned from Two Years as a Data Scientist

#51

Post by rejectedandsad »

chriak8292 wrote: post #39
cobertos wrote: post #8

I really liked this post. Tons of small tidbits I feel I'd only get from working in the teams they worked on. Some things I noted: * Check out `ray` as an alternative to `multiprocessing` * Check out `tqdm` * Use `pdb` more * See if fast.ai or https://jalammar.github.io/illustrated-transformer/ are worthwhile * Prioritize the papers I read better * _Leveraged_ index funds?

Risk-adjusted return (Sharpe) - VTI/VOO: 0.6 to 0.8 - Leveraged S&P500 ETFs: 0.5 to 0.9 - Savings account: positive infinity - Treasury bonds: 1.0 to 1.2 Leveraged index funds give approximately the same risk-adjusted return as passive index funds but with 2-2.5x the absolute return (eg, UPRO). However, if you have a weak stomach (you don’t like seeing your balance drop), then leveraged ETFs are not for you.

What should the net worth of someone like OP be by 25? I really wonder how far behind I am compared to my more savvy investor peers.

Re: Lessons Learned from Two Years as a Data Scientist

#52

Post by nerdponx »

joeman1000 wrote: post #42

Am I reading this correctly?: you were hired at MS with only cursory python knowledge? I’m very jealous and thinking more and more about going into IT after I leave uni for engineering (not software engineering). I know python well, along with a few other things I’ve picked up over the years (emacs/elisp, vim/vimscript, LaTeX formatting, JS, Common Lisp, APL, bash scripting, mathematica and matlab etc.). Would this b…

He was an academic researcher.

Re: Lessons Learned from Two Years as a Data Scientist

#53

Post by nerdponx »

Zababa wrote: post #18
jupin wrote: post #12

I love the tip about using the python debugger "pdb". This reminds me of the similar Ansible debug feature (e.g. "debugger: on_failed") which let's you jump into an in-flight playbook.

I remember using OCaml a bit and it has a time-travelling debugger. You launch the program with the debugger, it crashes and then you can inspect the program as you wish. I was really impressed by this.

Someone wrote a time-travel debugger for Python, but it seems like a one-off project and I'm not sure if there's an actively maintained/developed tool. https://github.com/TomOnTime/timetravelpdb

Re: Lessons Learned from Two Years as a Data Scientist

#54

Post by ZephyrBlu »

rejectedandsad wrote: post #51
chriak8292 wrote: post #39

Earlier quoted context omitted.

Risk-adjusted return (Sharpe) - VTI/VOO: 0.6 to 0.8 - Leveraged S&P500 ETFs: 0.5 to 0.9 - Savings account: positive infinity - Treasury bonds: 1.0 to 1.2 Leveraged index funds give approximately the same risk-adjusted return as passive index funds but with 2-2.5x the absolute return (eg, UPRO). However, if you have a weak stomach (you don’t like seeing your balance drop), then leveraged ETFs are not for you.

What should the net worth of someone like OP be by 25? I really wonder how far behind I am compared to my more savvy investor peers.

Comparing net worth when you're young makes no sense. It only makes sense in 20 years after things have had time to play out, and even then you still have a lot of time left.

Even if you have a far superior strategy and an equal amount of starting capital you might not be that far ahead of your peers after only a few years, and even if you are there is no guarantee your strategy will continue working.

In reality the amount of starting capital, income and investment strategy will vary wildly among your peers so almost any comparison is a wild extrapolation.

FWIW I'm younger than you and still have negative net worth (Loans), but growth rate > net worth. I have many years left to earn money, invest it and let it compound.

Re: Lessons Learned from Two Years as a Data Scientist

#55

Post by ineedasername »

lordgrenville wrote: post #26

This feels like someone's private rough notes, but seeing as it's on the front page I have one nit-pick: > OS packages...which are installed with apt-get (linux) or homebrew (mac) (Home)brew isn't bundled with Mac, and also works on Linux. And lots of Linux distros don't use apt.

And lots of Linux distros don't use apt.

We don't like to talk about those. It's just not done in polite company. Like that distro that concerns itself with hats. RPMs should apply to cars, never packages.

Re: Lessons Learned from Two Years as a Data Scientist

#56

Post by joeman1000 »

nerdponx wrote: post #52
joeman1000 wrote: post #42

Am I reading this correctly?: you were hired at MS with only cursory python knowledge? I’m very jealous and thinking more and more about going into IT after I leave uni for engineering (not software engineering). I know python well, along with a few other things I’ve picked up over the years (emacs/elisp, vim/vimscript, LaTeX formatting, JS, Common Lisp, APL, bash scripting, mathematica and matlab etc.). Would this b…

He was an academic researcher.

That makes more sense. Thank you.

Re: Lessons Learned from Two Years as a Data Scientist

#59

Post by natchy »

MontyCarloHall wrote: post #44
qsort wrote: post #4

> Google doesn't allow any production-level projects to be written in python due to safety concerns Is this actually true? If it's true that Google doesn't allow Python in production, it seems unlikely that it's due to security concerns.

Their command line interfaces to Google Cloud Platform (e.g. gcloud/gsutil) is 100% Python. Is that not considered a “production level project”?

That’s just a client though, right? They write JavaScript too for clients.

It’s probably production level services, not just projects.

Re: Lessons Learned from Two Years as a Data Scientist

#60

Post by ElevenPhonons »

It's worth noting that "except:" is not the same as "except Exception:" in Python. "except:" is catching BaseException which is often not what to do. BaseException is catching SystemExit amongst other things.

- https://docs.python.org/3/library/exceptions.html#exception-...

Post reply on HN