I gave a talk at PyAmsterdam today and it was a lovely community. I get the chance to answer some questions that have been puzzling me for a while. I ask people to vote for me on DirectPoll (first time trying it) so I know what the community thinks.
It is a questions that I have been thinking since I was a naive Python Data Scientist. “Why I can’t just do it in a for loop?” came through my mind all the time. Why we have to follow certain convention in coding? Is Pythonic a thing or just peer pressure.
Almost 90% of you things that it really is a thing. (35 votes)
For me, I am not a fan of SQL. Date back to my first data science job I was furious about writing thousand links of SQL just to get some aggregated results. Joining tables are not fun as mistakes can be made easily. As we are designing a new query language in TerminusDB, I want to know what people things about SQL.
I am surprised that 70% of you like SQL! Hmmmmm… (38 votes)
During the time I was translating WOQLjs to WOQLpy I wonder what how shall I make the query building more “Pythonic”. What would Pythonistas prefer? Chainable calls like
WOQLQuery().doctype("journey").label("Journey) or Pandas DataFrame style, multi-parameters calls like
WOQLQuery().doctype(id="journey, label="Journey"). (I failed to show the result in the talk so here you go!)
Since Pandas, the most popular data manipulation library in Python uses the multi-parameters calls, I am not surprised that 80% of you would prefer that. (25 votes)