chartpath

9 ms·

61.
by chartpath 5y ago
There are also use cases for dynamically generating code, or having higher kinded types that Go just can't touch. Anything that resembles data engineering is unnecessarily tedious. One of my favourite tools for teams these days is Hasu
62.
by chartpath 5y ago
18 years experience here. I have recently converted to startup employee (again), giving up a lot of money in the contract market exclusively because I am motivated by the business case and have a deep respect for one of the founders, who is
63.
by chartpath 5y ago
I've been dabbling in expert systems (aka rules engines) for the last few years. See CLIPs and Rete algo in your language of choice. This, combined with exposure to KNN over TF-IDF (either live indexes like Elasticsearch under the hood
64.
by chartpath 5y ago
OMG as a long-time Pyramid user I never thought I would see an article referring to the Zope Component Architecture on the front page of HN.
65.
by chartpath 5y ago
I know nothing of BTC but how would you stack it against gold, real estate, other traditional stores of value as a hedge against hyperinflation?
66.
by chartpath 5y ago
We live in a numpy array.
67.
by chartpath 5y ago
Agree with all the comments like "where are the nd arrays?" BUT, they have list comprehensions!! One of the main things I miss coming from Python.
68.
by chartpath 5y ago
Wasn't there an app on iOS way back at the start that was just an image of a ruby/gem of some kind that cost $999? Some people actually bought it before Apple took it off the App Store.
69.
by chartpath 6y ago
Love pg but lol at opening with "I write simple" and immediately throwing around some Italian words.
70.
by chartpath 6y ago
Interesting, thanks! Not an index per se but the Canadian Centre for Policy Alternatives publishes living wage reports, here is how they calculated it in the last one (2019): https://www.policyalternatives.ca/sites/defa
71.
by chartpath 6y ago
Is there any way to replicate this for housing speculation?
72.
by chartpath 6y ago
What? Yes we do, see drone strikes [0]. As a paranoid citizen I don't trust people with my privacy if they get off on punching down at other people's freedoms. Simultaneously I agree that Eich built some interesting technology. Ju
73.
by chartpath 6y ago
https://techcrunch.com/2014/03/28/after-supporting-prop-8-br...
74.
by chartpath 6y ago
I thought the meaning of life was negentropy. We seem to habitually build little corners of order in the eddies of the waves of entropy sloshing around. Even just by observing things. If using entropy emission as the indicator, how would we
75.
by chartpath 6y ago
Thank you for bringing this up. It's outrageous that it isn't more known outside of Canada (and inside). The last one only closed in 1996. EDIT: and good on you for honouring her.
76.
by chartpath 6y ago
Yes exactly, I want to be able to do rich auditing of the predictions. Not sure about performance yet, still prototyping!
77.
by chartpath 6y ago
Graphs to organize annotation labels are necessary to avoid concept drift! Even though BLAS on graphs and fancy JVM property graphs are cool and interesting, my small brain is fully satisfied with LTREE labels in Postgres :)
78.
by chartpath 6y ago
That's just lack of leadership skills and possibly incompetence about subfields of development they don't have experience with. It can become toxic in its own way when the first bully/troll is hired since that person will pul
79.
by chartpath 6y ago
Currently https://github.com/nilp0inter/experta but https://github.com/noxdafox/clipspy seems nice, I just shied away from using it due to uneasiness about FFI and debugging, even though the origi
80.
by chartpath 6y ago
I still do this after 15 years! Commitment means creative autonomy though. If valid solutions get tossed out by impostors then it becomes impossible to sustain. Or when authoritarians and people low on agreeableness traits start trying to m
81.
by chartpath 6y ago
That's true, I'm making recommendations based on Multinomial Naive Bayes (and SGDClassifier) over custom TF-IDF bags of words, so it is like search plus text classification. And some endpoints do just check the cosine or Jaccard d
82.
by chartpath 6y ago
Yup
83.
by chartpath 6y ago
I've been using LASER from Facebook Research via https://github.com/yannvgn/laserembeddings to accept multi-lingual input in front of the the domain-specific models for recommendations and stuff (that are trained
84.
by chartpath 6y ago
True about Facebook. But didn't they completely re-invent themselves once the white suburban soccer moms took over, and then again when all the drunk racist uncles came on? I feel like you are right about the need for an original purpo
85.
by chartpath 6y ago
As the article says, that is not the kind of union they set up. They set up a "members only union" which is voluntary to join or not. Either you are unhappy with conditions and need protection so you join for the the support netwo
86.
by chartpath 6y ago
Not sure about actors but I just use joblib for parallelization. It is way lighter than Dask and friends but still depends on (cloud)pickle serialization.
87.
by chartpath 6y ago
I agree that premature abstraction is bad. Except when using a mature off-the-shelf tool, e.g. Keycloak. Sometimes if you know that you need to implement a standard and are not willing to put in the effort for an in-house solution, that lev
88.
Ask HN: How do you mentor developers?
2 points by chartpath 6y ago|1 comments
89.
by chartpath 6y ago
Yeah totally! I have done a couple different LTREE label columns on the same table before but that was for separating named entity recognition from other things, so still just different kinds of class membership (e.g. people, places, things
90.
by chartpath 6y ago
This is a really cool and very smart looking mathematical way of looking at something that I do with lower-tech taxon labels using LTREE in Postgres to store the annotations. In PG you can do a kind of KNN for things that are X LTREE degree

More ›