Wednesday, December 12, 2018

PsychoPy on Linux


We've been using PsychoPy in our lab for a few years now, developing mainly on Linux, but often collecting data on Windows (shared testing facilities). On the basis of this experience, I recommend installing PsychoPy on Linux thus:

sudo apt install psychopy

This gives you quite an out of date version (1.85.3, released Aug 2017, for Ubuntu 18.04, as of Dec 2018), but trying to build it from source, or with pip, is just a huge time sink, often fails, and can kinda mess up your system in the process.  

This leads to a further recommendation. For archive purposes, archive your .py file as well as your .psyexp file, as more recent versions of Builder produce .psyexp files that are not backwards compatible!  

In fact, for portability, seriously consider treating PsychoPy mainly as a Python library, and develop experiments entirely as .py files, i.e. avoid Builder entirely. Builder is a good teaching tool, but anyone who can write R scripts can also learn how to write experiments directly in Python.


Thursday, September 20, 2018

Plymouth University to open Brain Imaging Centre in 2020

Many years in the planning, yesterday we bought a 3T Siemens scanner to put in our new Brain Imaging Centre building (picture above). Due to open for business in Spring 2020.

Tuesday, September 4, 2018

Doom (2016) on Linux


https://proxy.duckduckgo.com/iu/?u=http%3A%2F%2Fwww.videogamesblogger.com%2Fwp-content%2Fuploads%2F2016%2F05%2Fdoom-2016-walkthrough-640x325.jpg&f=1

Valve recently announced Steam Play, a system that allows previously Windows-only games to be played on Linux. It's in beta at the moment, but I thought I'd give it a go with Doom (2016). After a bit of a faff, it seems to work pretty well. The faff was that, on initially running, you just get a black screen. A search through the forums revealed that ALT+Enter brings it into a window on the desktop (rather than fullscreen), and it then works. From there, you can press ESC to bring up the game options, and set it back to fullscreen. Played for about an hour - works pretty well!

Of course, Doom 2016  not open source ... but it's still interesting how good Linux is getting as a games platform.

Friday, July 13, 2018

Goodbye Hush Coffee


Hush Coffee, Plymouth -- where the Plymouth meetings of the Devon & Cornwall Linux User Group started, perhaps 20 years ago - just closed its doors for the last time. Very sad. Now on the hunt for a new home...

H-21

Shortly after leaving my h-teens, now 21. Yey.

Sunday, May 27, 2018

Second archival appearance


After posting that Acorn Programs link (see last post), I thought I'd also link to the first psychology publication I appeared in ... as a stimulus...

50 years of BASIC


This month marks the 50th anniversary of BASIC -- the programming language that taught me and millions of other people how to code. For me, BBC BASIC. I first made some money writing a BASIC program in 1984, publishing a listing in Acorn Programs.

Wednesday, May 23, 2018

Cargo cult writing

Richard Feynman coined the term cargo cult science, which can be defined thus:

"Cargo cults ... focus on obtaining the material wealth (the "cargo") of the advanced culture by imitating the actions they believe cause the appearance of cargo ... Similarly, although cargo cult sciences employ the trappings of the scientific method, like an airplane with no motor they fail to deliver anything of value."  https://en.wikipedia.org/wiki/Cargo_cult_science

Increasingly, I'm seeing in undergraduate student writing something that I think is best described as "cargo-cult writing". By this, I mean the construction of phrases that seem focussed on sounding scientific rather than communicating clearly. If you skim cargo-cult writing, it can appear scientific, but when you  read it for meaning, it is unclear, often ungrammatical, and says very little of substance.

Somehow, we need to focus students back on the idea that science is about communicating clearly, in the plainest English possible. Using complex, difficult-to-read language is sometimes a byproduct of talking about complex issues with precision -- but it's not a goal in itself!


Thursday, April 12, 2018

No longer an h-teen :-)


My h-index on Google Scholar just hit 20 - I'm no longer an h-index teen :-)

Supporting the R Foundation



Support the R Foundation by becoming an Contributing Member (25 EUR / year). R is a free and open-source environment with particular strengths in statistical analysis.

Friday, January 19, 2018

Top 1%

Quick follow-up on a metric I first reported five years ago - I'm still in the Top 30 categorization researchers worldwide (among about 4500 published authors). That puts me in the top 1%. Methodology here:  http://willslabblog.blogspot.co.uk/2013/05/top-30.html