Data analysis – Page 9

ScaleR package now available as part of free Microsoft R Client

July 12, 2016 By editor

The ScaleR package provides functions for performing scalable and extremely high performance data management, analysis, and visualization in R. It was only available to those who had a Microsoft R Server license—until now.

With the introduction of the free Microsoft R Client for Windows tool you can now work with the full set of ScaleR functions without having to part with a cent.

Of course, there’s a catch—there are constraints. Specifically

[…] the data to be processed must fit in local memory, and processing is limited up to two threads for ScaleR functions.

However, this allows you to prototype your analyses using the free client and push them to SQL Server or Hadoop using Microsoft R Server when you need to scale.

The problem with spreadsheets

May 9, 2016 By editor

A 2014 report by the Aberdeen Group includes the results of a survey asking what people found most challenging about spreadsheets. The biggest reported challenge was version control issues.

At least people are becoming aware of the problems—even if they are still failing to address them.

One statement in the report that depresses me is

…Best-in-Class organizations are 40% more likely than All Others to have enterprise software that has an interface that mirrors the look and feel of spreadsheets.

My guess is that this is correlation, not causality. But it’s indicative of a perspective that plagues the software industry. Most software is just a copy of something else. Very few people sit down and attempt to build efficient user interfaces.

There are off-the-shelf components for building grids and everyone knows how to use a spreadsheet, so let’s build something that looks like a spreadsheet. That’s how the argument goes.

However, a spreadsheet isn’t a particularly efficient interface. It benefits from the fact that we’ve all been trained in how it works, but this preexisting knowledge drags down true innovation.

Grids of numbers require interpretation. If I’m in front of a computer, I don’t want to have to think about numbers. That’s why I bought the computer. Software developers need to learn how to deliver insights—not data.

Designing a data strategy

May 2, 2016 By editor

Learning Tree just published my article on why it’s important to think carefully about the data and techniques employed in data-driven decision-making.

Throwing machine learning at existing data doesn’t cut it.

Introduction to Microsoft R Open and Microsoft R Server

May 1, 2016 By editor

Familiar with R and wondering what Microsoft R Open and Microsoft R Server bring to the table? Then check out Lixun Zhang’s article.

DeepMind has access to data on millions of NHS patients

May 1, 2016 By editor

DeepMind, the Google-owned AI company, has an agreement with the Royal Free NHS Trust that gives it access to healthcare data on 1.6 million patients who visit three London hospitals. The agreement includes access to five years of historical data.

The company is building an application to help monitor patients with kidney disease.

While it will be interesting to see what DeepMind can produce with such comprehensive data, the National Health Service sharing identifiable healthcare information with a commercial entity is sure to raise privacy concerns.