The Gray Lady hasn't faded some more, after all
Compared to the print edition, digital NYT has infinite news hole, so all the news that fits is no longer a constraint.
Data science
Dead simple bullseye map
Get a pattern working once by keeping it simple and then you can reduce scripts to produce different outputs by changing only a few lines.
Ducktyping
Wherein is compiled the argument against strict typing, with some slight acknowledge of the opposing views.
Julia has stolen my heart
Julia is a modern data science programming language that has clean syntax, the ease of use of an interpreted environment, like R, with the speed of a compiled script like C.
«Best practices» Convert the incoming hot mess
The wry joke is that data science is only 10% of the sexy stuff like machine learning—the other 90% is euphemistically called "data cleaning."
Specificity and recall—which is which, again?
Specificity and recall are two important metrics used in the field of machine learning and statistics, particularly in classification problems. They are used to evaluate the performance of a classification model, such as a binary classifier.
Form over substance
Sometimes "believe the science" is indistinguishable from "believe in your dreams"
Review of R Cookbook, 2nd Ed.
A recipe can encapsulate a cusine. This R Cookbook sneaks in the mindset needed to exploit the unsurpassed statistical functions of R.
Row collapsing
It’s not always possible to store everything in a tidy but hugmongous data store.
R: A Brief History
R has a decent pedigree for a domain specific language designed to be used interactively by statisticians and to accommodate new algorithms developed by statisticians. It was a monumental accomplishment also in light that it was the part-time work of a small core team.
The plural of data science is not formulae
Two mysteries: Why were the data not stratified into a training set and a test set? How could temporal auto-correlation fail to be addressed?
