elliott34

11 ms·

31.
by elliott34 12y ago
"a toolset that allows non- and minimally-technical employees to quickly and easily build new reports" Been there, done that. This is the mantra of the BI system marketing machine that I've never actually seen work in practic
32.
by elliott34 12y ago
But if you're setting up bokeh server for the first time...bring a friend and some coffee
33.
by elliott34 12y ago
Correct. That's why those businesses get the enterprise version and it's all hosted behind a firewall.
34.
by elliott34 12y ago
plotly is awesome. A lot of companies spending $50k + a year on BI software can do better (and save a ton of money) by throwing plotly iframes on some simple static webpage and update them at any interval using cron jobs and python.
35.
by elliott34 12y ago
Does the author have something backwards or do I? P(RateOfSuccess|Successes) = Beta(α,β) "In Bayesian terms, C3PO's estimate of the true rate of success given observed data is referred to as the likelihood." BUT we know likel
36.
by elliott34 12y ago
It's only CA at this point.
37.
by elliott34 12y ago
Car insurance for the major players in the U.S. has historically been about capturing users in the funnel, and making pricing information extremely opaque. This is very different from how the UK operates (see https://www.google.c
38.
by elliott34 12y ago
Checkout BigML.com, it has a ton of visualization As far as creating a model and deploying it to production, I cannot think of anything simpler and more wonderful to use than bigml. Outputs the decision trees /randomforests (via python
39.
by elliott34 12y ago
This is fantastically well written but goes over my head in a major way. Where can I read about the 101 version of thi article
40.
by elliott34 12y ago
This is crazy and awesome. Can't wait to see what happens.
41.
by elliott34 12y ago
Like most apartment/condo city dwellers I can't receive packages unless I am home. I would pay 30/month for some kind of external lockbox or some kind of capability to order anything I wanted to that box (outside my apartment
42.
by elliott34 12y ago
The psychological benefit of paying off my student loans early was immense. While mathematically illogical and not in my best interest, I sleep better at night knowing that if I wanted to drop everything I was doing and do something else, I
43.
by elliott34 12y ago
Now if we can only get sublime text 2 within an ipython notebook on the cloud..... For me, R was my first language, and then I learned Python, and beautiful things like list comprehensions, and it just clicks with my brain a bit more. In pa
44.
by elliott34 12y ago
Great commentary. Encoding high-level categorical variables as unordered integers, or even frequency count maps, is pretty common practice, but really only valid (as the author says) in tree based methods. Given deep enough trees (or enough
45.
by elliott34 12y ago
I love pandas much more than R but GOD I love Rstudio. Such a great IDE. Rstudio server, actually. My equivalent is running ipython notebook on an ec2 instance, which....is fine, but is a lot of scrolling.
46.
by elliott34 12y ago
Excuse my ignorance here: so are these functional programming paradigms?
47.
by elliott34 12y ago
Definitely see what you're saying with the prepare_data1()...prepare_data(2). Maybe a more functional approach would be suitable. At the end of the day, data goes into a tunnel and comes out in a different form, into a machine learning
48.
by elliott34 12y ago
Sure! I'd love to take a look. I've actually written a mysqldb library in python that makes "data objects" For example mydata=dataobject(query, connectionstring()) mydata.query() # get data mydata.data # pandas data fram
49.
Ask HN: I am a data analyst and my code is a mess
5 points by elliott34 12y ago|11 comments
50.
by elliott34 12y ago
Do you have any links to articles on the current benchmarks that you mention? Not being snide just curious to read more.
51.
by elliott34 12y ago
Considering enterprise BI software/platforms these days can costs upwards of 4k per month, this is roundoff error.
52.
by elliott34 12y ago
we've shifted to plot.ly for visualization and moved away from expensive BI tools. We set up python/pandas scripts on CRON to output realtime data to our local plot.ly web server, which makes a local copy of the data and updates
53.
by elliott34 12y ago
OOOOOO awesome. Thank you!
54.
by elliott34 12y ago
I work in pandas 95% of my day doing data automation tasks, manipulating sql queries, and data-munging for machine learning. It is literally life changing for someone like me. I used to program solely in R, but after discovering pandas I re
55.
by elliott34 12y ago
Why is that panda's problem? If you have scipy and numpy installed, it's just sudo pip install pandas....
56.
by elliott34 12y ago
Do you work from home?
57.
by elliott34 12y ago
R is an alien ray gun, full of crazy features but all the buttons are in a bizarre language and require 10 fingers
58.
by elliott34 12y ago
I have hashimotos and celiac disease. So there's n=1. The worst part about having celiac disease is that I do not have any symptoms whatsoever but am forced to eat this way so that ten years down the road, maybe kinda sorta, I wont get
59.
by elliott34 12y ago
This pairs well with the recent AirBNB fraud detection article. Airbnb has a lot of people to construct a robust machine learning architecture. For small start ups trying to make a recommendation engine, why not do sciencebox+ yhat cloud&#x
60.
by elliott34 12y ago
does anyone know if Java can port a gradient boosting model from R

More ›