Differential privacy is an area that makes some guarantees about not letting personal information about individuals escape. Might be a useful technique as well.
We've tried using docker to put R applications into production. I'm no docker expert, and I found it pretty easy to do.
A couple of pointers:
- If you have a dependency on a library like blast that takes a long time to compile, you can make a base docker image that already has that library installed. That makes iteration quicker, as you'll only need to build that base image once.
- If you put a web interface on the image using shiny, then it is straightforward to deploy it for your users to interact with.
We usually dockerize outputs of our Syberia projects as well. We have several dozen internal packages consumed by the root projects. With many contributors working on constituent packages daily, we've found frequent changes to packages can slow down a docker-only workflow. So far, using a base docker image with lockbox catching us up to the most recent daily and hourly changes has been working well.
That is exactly the method they use. Calculate the centre of economic gravity, then project it onto the closest point on the surface of the globe.
Here is a nice analysis that does the same process in a more intuitive way. It calculates the centre of economic gravity in 2D map coordinates. This gives results that are easier to understand, because everyone's used to looking at 2D maps.
It might be better for the upper map to select a colour scale that has a zero point on it - this would represent countries that have the same price as the dollar.
A diverging colour scale would make it easier to distinguish which countries are cheaper and which are more expensive. For example:
I actually had no clue how to do the color scale, and I ended up doing it with 5 different colors based on quantile.
Did a small write up with my thoughts about it: http://mts.io/2014/05/07/spotify-pricing-index/
Note that this doesn't apply retrospectively to your data, so it will only apply to data collected after you make a change. You might want to create an advanced segment which filters out these domains, or only includes the ones that you want.
Yes, but when people refer to the U.K. as a region, they usually mean the U.K. (i.e. including Northern Ireland) and the Republic of Ireland. Ireland is seldom its own region.
There is no mention of the Republic of Ireland in their earnings report, yet their European headquarters are in Dublin. "U.K." in the report means the U.K. and Ireland.
Really? Where have you seen this? And what evidence do you have that Google's report isn't including Ireland in its "Rest of World" figures?
To include an entire country under the heading of another neighbouring country without making that explicit is a mistake, not an assumption, doubly so when Google have quite large (and distinct) corporate presences in both Ireland and the UK. It wouldn't be too far from Starbucks listing all Canadian and Mexican profits under "United States" when there's a perfectly normal term to describe the geographic region ("North America"), just as there is with the British Isles.
https://en.wikipedia.org/wiki/Differential_privacy
Agree that strong guarantees about privacy aren't achievable.