This is cool!
I did find that using the station name is unreliable. Also, it’s hard to use a geospatial query using the new bigquery capabilities, because boundaries of cities do not have clear shapes (like a circle or a rectangle).
Awesome write-up, thanks for sharing!
The point about the Foreigner Team also existed in companies I worked for when living in Amsterdam, where there are special programs for high-tech companies to grant visas to engineers from across the globe.
Wow. I am totally blown away by both your transparency levels and scientific approach. This is one of the best and most meaningful posts that I have read on Medium, so thank you for that.
My feedback regarding the general approach (you really do so many thing in different areas so I think it would be easier for me to express…
Awesome! I wasn’t familiar with gophernotes.
For “classic” data science tasks we’re using Python, Go is used for data engineering parts at the moment (fetching data, some transformations, etc.). Python is kind of THE Swiss knife today when it comes to data
Ori Avraham thanks!
Regarding the first question: I would assume using a number high than
runtime.NumCPU()will not be very effective, but this should be benchmarked to make sure.
Regarding using a working pool, unless I’m missing something, it’s just another approach to control the levels of…