Live data from Hacker News

Lessons Learned from Two Years as a Data Scientist

dawndrain.github.io

41–50 of 89 posts

Re: Lessons Learned from Two Years as a Data Scientist

#41
post #38

Earlier quoted context omitted.

SPXL also has an expense ratio over 1%, which will eat away at earnings unless in the best of bull rushes (now).

Leveraged EFTs outperform VTI/VOO (in terms of total return) over 30-40 year investment horizons. Period. Now, the risk (potential one-year downside) is not for everyone.

A little bit of leverage as others have commented is fine.

The problem happens if there’s 51% drop in 2x levered fund.

There’s a reason the fund the article’s OP is in started in 2008 and not 40 years ago.

Re: Lessons Learned from Two Years as a Data Scientist

#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 be enough to land a position like yours? I am lacking in the AI area, but I can begin that on my weekends.

Re: Lessons Learned from Two Years as a Data Scientist

#44
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”?

Re: Lessons Learned from Two Years as a Data Scientist

#45

Earlier quoted context omitted.

> did you do any DD other than taking his advice? I did some backtesting simulations that made leveraged investing look pretty awesome. The effective borrow rate for funds like spxl is crazy low, way better than if I were to borrow myself. (Also, fwiw I was pretty conservative and am overall only around 2x-leveraged.) The internet is very opposed to leveraged investing imo, but I think most of the concerns are pretty…

Thanks for you perspective. I'm considering using leverage so it's interesting to hear from people who are currently using it. Also, you might find this tweet and paper interesting: - Tweet: https://twitter.com/patio11/status/1432891941138563077 - Paper: https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.89...

Use keyword “HedgeFundie” (username that first started the discussion on Bogleheads and now most refer this strategy with his name) to search discussions about Leveraged ETFs on Bogleheads forum and Reddit. There are over 300+ pages worth of discussion on this topic only on Bogleheads.

Also checkout information about permanent portfolio put forward by Bridgewater Associates.

Re: Lessons Learned from Two Years as a Data Scientist

#46
post #16

About Java vs Python, modern Java would be: import java.util.ArrayList; var cars = new ArrayList (); Python would be: cars: list[string] = [] The big difference seems to be that ArrayList is not a "default" data structure in Java, but it is in Python. While I like the Python example better, I'm not offended by the modern Java.

That modern Python example scream regression to me. Why not simply cars = []? This obsession with killing perfectly good languages with strongly typed hints is completely undermining the point.

I have work with large python project without types. It's a nightmare. Types are extremelly useful.

Re: Lessons Learned from Two Years as a Data Scientist

#47
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.

I heard this second-hand, not totally sure it's true

It's not.

The remote connection to safety (did you mean security?) would be that static source analysis tools don't work as reliably with dynamic languages. That matters at Google. But you don't even have to think as hard about it: Python is simply comparatively slow and inefficient. Google's fleet is large. It pays off to use more efficient languages.

(There's also the whole thing about Python being largely single threaded and computers being very wide these days, as well as being a terrible memory hog and memory making up half the cost of servers.)

Re: Lessons Learned from Two Years as a Data Scientist

#48

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…

What’s wrong with my comment? I’m just jealous of you IT guys..

Re: Lessons Learned from Two Years as a Data Scientist

#49
post #27
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?

Check out ipdb instead of pdb. It’s pdb but with the ipython repl instead of python’s

I find that PDB++ (on Pip as "pdbpp") is better in its role specifically as a debugger.

Re: Lessons Learned from Two Years as a Data Scientist

#50
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.

Yeah, did they rewrite YouTube?
Post reply on HN