Posts

I believe writing notes is essential for scientific research. Both for future reference, and also to provide space to think. I develop a lot of my understanding on paper. Once I’ve read and thought a bit about a topic, I really like starting with a fresh A4 page (or double spread) and writing down pointers to everything I know about the subject, little diagrams and snippets of mathematics. This would be wholly useless for …
3 min read

A note on Schultz In Schultz1959 (https://doi.org/10.1103%2Fphysrev.116.526), a Feynman polaron radius is defined from the Gaussian form of the Simple Harmonic Oscillator specified by the Feynman solution to the polaron problem. I rederived some of these results (and the key thing - figured out that a 0.44 ‘magic number’ was actually 2/9, i.e. 0.4 recurring), and decided to put the info out there somewhere. The core …
2 min read maths

ARCHER pulls 1.2 MW [1] to run it’s 4920[1] compute nodes (each of which is 24 cores). That’s just under 250 W power per node: 6 kWhr of energy in 24 hrs. A 24-hour single-node job consumes 8.6 ‘kAU’ in the internal compute units used for accounting. So that’s roughly 0.7 kWhr / kAU, or 700 kWhr / MAU. The UK electricity grid carbon intensity is roughly 400 g / kWhr averaged over the year: making 280 …
1 min read computing

Open-source software licenses are intensely boring things to read about. The choice available makes it very easy to default to not organising a license. Particularly when you are a serial procrastinator like me! However I would love for others to use my codes. It’s why I wrote them. So I sat down for a concerted half hour reading. My personal bias having benefited so much from GNU software would be for the GPL.
1 min read licensing

Starting my PhD in 2007, I spent many many years in seminar and conference talks on Photovoltaics being confused by the enormous array of experimental probes for photovoltaic materials. This is an attempt, stimulated (as always!) by questions from a student, to put down a list of standard experimental photovoltaic (or other optoelectronic) device characterisations. High energy theorists often ape Gell-Mann in describing solid-state …
4 min read

A little update to the book club. It’s going well! We’ve decided to work our way through John M. Ziman’s Principles of the Theory of Solids (2nd Edition). I really like this text, as I find the style very agreeable. It’s very conversational, in it’s best bits it feels like standing at a black-board having something explained just to you, the reader. § 1 introduces periodicity as the central theme …
3 min read Bookclub

Having your publication history in a LaTeX document is not the worst idea in the world. Updating a .bib file is natural and easy when making incremental changes (for the next, seemingly inevitable, academic application in 6 months time). A recommendation though—don’t decide to do this two hours before a submission deadline! 1) Get the data. I only discovered this mass-export by accident. Go to your Google Scholar page …
2 min read

Now that the Walsh group has a critical mass at Imperial College London, I am planning to restart the Bookclub. Keith Butler kicked these off in Bath with the guiding words of everyone’s favourite German pessimist, Arthur Schopenhauer: “Buying books would be a good thing if one could also buy the time to read them in: but as a rule the purchase of books is mistaken for the appropriation of their contents.
2 min read Bookclub

Summary: If you want to do a geometry relaxation around a constrained degree of freedom (bond length, angle, dihedral, etc.) in Gaussian 09, you need two ‘ModRedundant’ specifications, one to Build in the coordinate with the value you set, the next to then Freeze this coordinate so that it is constrained during the optimisation. Here is an example for Biphenyl, setting the torsional angle to 90 degrees (making the …
4 min read

If we consider the Mexican Hat potential discussed in the previous blog post, as a stand in for a soft phonon mode, a fundamental question in chemical physics is how regularly the mode tunnels between the vibrational states localised on one side, versus the other. Classically, the chance of crossing over is zero, unless you have sufficient energy to overtop the barrier. From a Statistical Physics point of view, at finite …
4 min read

I’ve spent a lot of time over the last three years thinking about electron-phonon coupling. Thinking, not doing, being the operative word, as I’ve found the theory really quite a large mass to get my head around! Recently I saw my way in to something that seemed vaguely useful and interesting. I’d put together a 1D Time-Independent-Schrödinger-Equation solver, originally for my own amusement, then to show my …
7 min read

I applied for the Royal Society University Research Fellowship scheme last summer. I wasn’t successful (Boo). As part of the application, I wrote a lay-person summary of my research project, limited to 3500 characters. There wasn’t much guidance as to who or what it was for, so I wrote it for a typical imagined audience of A-level students. I had quite a lot of fun writing it (certainly more fun that trying to get the …
3 min read

Mark Tuckerman (MET) visited London on the 31st March 2016, and gave a highlight seminar to the TYC at UCL. I had been really enjoying his recent (2010) Computational Stat Mech book [1]. I know very little about structure prediction, but it is of general and captivating interest, and an essential step in truly predictive computational materials science. Here are a summary of my notes from his talk, written up mainly for my own …
8 min read

All modern functional electronic devices are made from semiconductors. A semiconductor is a material with a moderate band gap. But to turn this semiconductor into a device, you need to be able to selectively inject and extract electrons from above and below this band gap, forming selective contacts. The metrics we want to calculate are the Electron Affinity (energy gain by adding an electron from infinite distance) and the …
7 min read

This blog used to be on wordpress, at https://jarvistmoorefrost.wordpress.com. ( There are a few comments by other people scattered amongst those posts, I’m not sure how to best transfer / archive these. ) Files were extracted to a plain text Markdown format, and are now published with Hugo with the nofancy theme (slightly tweaked). I moved this blog from wordpress as I had grown irritated with the editing experience, and I …
1 min read

Band Theory of Solids: An Introduction from the Point of View of Symmetry Simon L. Altmann Clarendon Press, 1994 This is a fairly esoteric publication. The stated intent was to provide an introductory (I guess suitable for an undergraduate course) book on solid state theory, that none the less contained a strong dose of group theory from the get go. This is an interesting idea, but I’d say that it’s only partially …
4 min read

I’ve learnt more from John Ziman’s books on electronic structure, phonons, and disorder, than from any other single source. This is a little guide to his academic works. 1979 - “Models of disorder”. Though his most recent book, from the printing you would assume it was from the 50s! This is the only real attempt to write a textbook summarising all approaches to disorder in electronic structure modelling. As …
3 min read

There are a number of groups at Bath University doing work of interest to Condensed Matter Theorists. This is just a quick set of ‘hot links’ to the individual seminar sites + pages. It is not intended to be exhaustive or complete, but I try and see what’s on at all of these events for every week. Physics: Theory group seminars: http://people.bath.ac.uk/rj248/theory_seminars.html Maths: General colloquium, …
1 min read

I gave up OPENMP’ing my Monte Carlo code about a year ago. I was trying to do it at the level of MC moves, and there are massive issues here in that the different threads are trying to sample and update the same simulation space - leading to segfaults as the variables crash into one another. One method I’ve seen used in the literature is domain decomposition - where each thread runs MC on a segment of the simulation …
2 min read

The workflow in computational materials modelling is often highly complex, and bespoke to the particular calculations you are doing. The flow is highly derivative (results of later jobs depend on work on prior jobs), but can also be quite chaotic and multiply branching. Unsurprisingly, your files easily end up very messy, with significant confusion about what steps were required to get to the present working setup. Version control …
2 min read

(Updated 2016-03-12 to re-add the links) I am probably the last person you would want to consult on tips to write a scientific publication. I’m slow, I procrastinate, my English is not the best, much work lies on the cutting room floor - skeletons of half assembled papers. Still, ‘that which we are, we are’, though my skills are lacking I’m very interested in improving, and certainly enjoy reading a well …
5 min read

Solid state materials pack together in a myriad of ways. The crystal structures are typically referred to with the name of the archetype compound. This ranges from fairly common English words such as ‘diamond’ and ‘rock-salt’ to, ‘zinc-blend’ ‘wurtzite’ and, of course, ‘perovskite’. The atomic packing within these structures is fully described by their space-group, of …
2 min read

I polished some slides I gave at our group meeting last summer, on using ‘ssh’ proficiently at Bath university. I thought they may be of use to other’s, so published them online. They mainly cover the setting up of a ssh-config (the examples are probably useful for most people that use HPC facilities), and tunneling via the LCPU facility (fairly Bath specific, though other academic institutes might offer a similar …
1 min read

1 min read

I made a scientific poster last week, for the first time in five years. I drew it in Inkscape, a free opensource vector drawing package. I’m have quite a lot of experience using Inkscape in drawing cave surveys, but I believe it’s both relatively easy to use, and has some extremely useful tools for constructing posters (most notably the alignment & distribution tool). The one gotcha is that you have to explicitly …
2 min read

So I was live-plotting some results from a simulation code (written in C), which was producing results for an input parameter (in this case Temperature), over a range generated by a for loop. Pretty standard computational physics. It was clearly suboptimal watching the new dataset arrive in small, fine, increments from zero upwards; the obvious thing was to change the discretisation of the for loop & make it more coarse. Yet, …
2 min read

This looks like it might be extremely useful for a personal knowledge base / article discussion. It’s setup more for humanities work currently (books, biographies and authors), than science (some way to import a DOI to build a paper-discussion place would be fantastic). A collaborative journal club / papercore / paper summary might also work wonderfully within here. http://tw5.scholars.tiddlyspot.com/ Update 2017-01-01: …
1 min read

1 min read

Julia 0.3 Release Candidate 1 has quietly appeared on the nightly builds page; release notes in the very active GitHub repository - improvements to the REPL (tab conversion of ‘latex maths’ to unicode), and a lot of interesting linear algebra improvements. The inaugural JuliaCon has been held, and the presentation slidedecks are already online: https://github.com/JuliaCon/presentations; videos are coming but the upload …
2 min read

This is a collection of bookmarks I’ve collected over the years of useful online physics notes; hopefully this will be of some use to others. My bias is as a computational physicist doing electronic structure theory for materials modelling, so all within the remit of condensed matter theory (no high energy physics, for instance, and nothing that would count as ‘theoretical’ research for the last forty years at …
3 min read

I’ve regularly dallied with keeping a computer / online logbook. I’ve tried many products, even wrote my own (retrospectively crazy) set of scripts called ‘freck’ (https://github.com/jarvist/freck). All these platforms were only ever used for a week or two before falling by the wayside. I also don’t like the idea of walled-gardens (such as evernote) becoming a trap for my information. I keep paper …
3 min read

These days I almost exclusively make my talks in Google Drive / Present. It just works, I can work on my talk from anywhere (my laptop is Linux, my work machine is Mac OS) the modern interface is really quite nice (at least in Chrome). Most importantly for me it has very sensible auto-hints for placing items (a feint alignment lines pop up when you are centre / left / right aligned with other elements on the page).
2 min read

1 min read

1 min read

Day of the deadline & you realise you need to convert all your figures to EPS format, and put them in the same directory as the source .TeX files for submission? Happened to me today with APLMater. I had 23 individual files (as the figures were full-width composites), so I wrote a script to strip the filename from the TeX, rename to a flat file format, strip any periods (‘.’) in the filename & do the conversion …
1 min read

1 min read

In science we are often rediscovering the already known, this happens on a personal level almost every day. You feel dumb and insecure, surrounded by the condensed wisdom of lifetimes in science, textbooks of seemingly impenetrable theories whose every page requires hours of study to have the merest glimpse of the knowledge within. Recently I’ve been trying to think about how to represent the alignment of a stick within a …
2 min read

I came across the following blog post while reading twitter before bed (perhaps a mistake): http://sylvaindeville.net/2014/03/24/the-making-of-a-paper-a-praise-for-slow-science It really irked me. The general description of scientific discovery was extremely interesting. Even the self-justifying air of ‘slow cooked’ science, and not ‘salami slicing’ his papers (as perhaps he feel’s critical colleagues …
3 min read

1 min read

I’ve just bought a new (secondhand) Thinkpad. My previous, a X60s (bought for £150 in Oct 2009) with a cheap generic 8-cell battery, has done me well over the years. Last year I upgraded it to a Samsung SSD, which made playing on the command line super speedy, but websites tended to drag a bit, and PDFs sometimes staggered between frames for tenths of a second when giving talks. The main issue was that the lovely matt 4:3 …
3 min read

1 min read

1 min read

Today is the launch of what I believe is the first organic solar cell MOOC (Massive Open Online Course). It’s being delivered by Krebs, Bundgaard and Jørgensen from DTU. The syllabus covers life cycle analysis and characterisation techniques so I think it would be useful for anyone starting research in Photo Voltaics in general. Organic Solar Cells - Theory and Practice - Coursera Most exciting for me is that you can order an …
1 min read

In academia you often want to keep code in a private repository (such as a free BitBucket repo), then push out to GitHub for greater opensource exposure when you’ve mostly finished the first paper etc. This is easy to do with Git while retaining full commitment history by adding an extra remote. I just did this today, and leave these simple commands here as a bookmark for when I next need to do it.
1 min read

Sharelatex is now opensource: https://www.sharelatex.com/blog/2014/02/21/sharelatex-is-now-open-source.html; discussion on HN: https://news.ycombinator.com/item?id=7276263. They also now offer a ‘personal’ account which allows for unlimited projects + one collaborator. The other big player in this area of collaborative online Latex document editing is https://www.writelatex.com/. I’ve generally found their web app …
1 min read

Our fairly prospective paper is out to review with Nano Letters at the moment, the preprint has just gone live on the arXiv. There’s a massive range of calculations presented in quite a short paper - from GSGW (MvS), periodic plane wave DFT (FB) & CCSD quantum chemical calculations (JMF) to classical electrostatic calculations (JMF). We make suggestions as the to the likely effect of these results on device operation, and …
2 min read

1 min read

1 min read

My one new year’s resolution is to use DVCS everywhere. With free BitBucket Academic accounts (unlimited private / public repos), http://blog.bitbucket.org/2012/08/20/bitbucket-academic/, there’s no reason not to use it for every paper, internal document, code snippet or website. From a Linux / OS X console there is very little friction in automatically using “git add + commit, push & pull”. It’s …
1 min read

1 min read

I must admit that I have a lot of ‘historical’ (as in, before the last hour…) directories with processing scripts named usefully ‘process_processed_data.sh’ and such. If I remember, I often pipe a ‘history’ into a hist.sh file before I close the terminal + forget what I’ve done and how to process the data. So I definitely should and could do better, so why not use Make? It handles …
1 min read

ARCHER is the new EPSRC/NERC High Performance Computing facility, the successor to HeCToR. We have early access to it during its testing phase (until 16th Dec). The transition has been extremely impressive, the new machine is stunningly powerful (#19 in the world on the Top500 Nov 2013 list, a Cray XC30 on just-fabbed Intel Ivy Bridge processors ) and everything seems to more or less just work. The login nodes are responsive, the …
1 min read

Mavericks and Failed ARP causing network drops - solved Since I started at Bath on an OS X Mavericks machine, my SSH sessions were pausing for 4-6s every minute or so (randomly). Playing with Ping, I saw that correct packet routing was simply failing (averaging 7% packet loss for ping). Much wailing and gnashing of teeth later, and I found a simple fix. OS X Mavericks Unicast ARP packets were failing - so turn them off with …
1 min read

Advanced R Programming Advanced R Programming is being written in an open manner by Hadley Wickham (ggplot2). It’s nearing completion & printing, but the Wiki form of the content will remain online in perpetuity. R is a really difficult language to google answers to (it’s all short-hand functions & punctuation <- $ ^ which gets stripped or is not very searchable), and the documentation I’ve found so far …
1 min read

Coming to Mac OS X from over a decade on Linux is quite a change! Here are notes of things I’ve found useful in making the transition, starting with more general and ending at things rather specific (and esoteric!). Homebrew is sort of a Gentoo ports - http://brew.sh/, it’s pretty swish and pulls directly down from GitHub based ‘recipes’. There’s a ‘science’ ‘tap’ (PPA) with …
2 min read

From the start of November 2013 I’ve moved down from Imperial College London to Bath University, joining as a PostDoc on an EPSRC Programme Grant (Energy Materials: Computational Solutions). Here, I am in the Walsh Materials Design Group - still PhotoVoltaic based, but more concerned with Perovskites and Kesterites than Fullerenes. Crystal structures have the most fantastic names - after the Georgian mineralogists who …
1 min read

The first book that I bought during my PhD was the excellent (and cheap!) Szabo and Ostlund ‘Modern Quantum Chemistry’. The development of the SCF method and then treatment of MBPT (which is where the beautiful 4th order Goldstone diagram on the front cover comes from) is absolutely fantastic. However, Szabo is a bit too dry to dig back into for a quick reminder. I think, unfortunately, that graduate level QM/QC is a …
2 min read

1 min read

1 min read

I was moaning to a PhD student in the group about the perennial issue of checking representative torsional conformers of oligomers. In this case it was the question about whether you’d see a large variation in oscillator strength (of a TDDFT calculation) if you change the conformation (we don’t see a big variation in the energies of excited states, unless you break conjugation). This is a genuine issue as the potential …
4 min read

1 min read

1 min read

I finally got around to (minimally) cleaning up my code from the 2006 Nanoletters paper and put it up on github (https://github.com/jarvist/Amphisbaena). This simulates coarse-grain on-lattice morphology generation of polymers via a slithering snakes / reptation model [amphisbaena], and then also enables you to carry out mobility simulations via Time of Flight (ToF), drift velocity or diffusion [gorgophone]. It’s coded in C …
1 min read

YouTube: NeuroTrends: Large-scale automated analysis of the neuroimaging literature; SciPy 2013 Presentation Interesting presentation on doing meta analysis of results in scientific papers via automated scraping with a python code. For me the ‘sciscrape’ library is probably the most interesting thing (regexp based on the raw text): https://github.com/jmcarp/sciscrape Seems to have a typical story arc of developing an …
1 min read

Google Reader has finally shut up shop. Eight years, many scientific papers, snippets of technology news, new music and many many blogs; we had some good times. All the replacements currently available still seem to be a bit unsatisfying - Reader feeder updates were really quite amazingly fast, and the interface was clean. There will now be a lot of starred articles which I’ll never read, and old cached blog posts which have …
1 min read

On the 20th June, a team of mathematicians released a book on Homotopy Type Theory, written collaboratively over 6 months with GitHub and Latex. The content is of no direct interest to me - but the collaborative method of writing, and using git to work simultaneously on a large Latex project certainly is. Discussion of using git + latex on HN, including touching upon methods to be able to merge simultaneous edits (latexdiff etc.
1 min read

1 min read

1 min read

I often use the nice built in ‘ray’ raytracer in Pymol. These ray traces can take minutes to hours, and during that time I of course want to use my computer! Unfortunately, the Pymol window freezes during the ray trace, so when I switch virtual desktops it sometimes looks as if there’s a frozen terminal or similar there, which I click on, sometimes inadvertently, sometimes because I am a fool. This click gets …
1 min read

I’ve been struggling recently to get CDFT jobs to converge in NWCHEM. Rather than continue to play with ~100 heavy-atom structures (of relevance to organic electronic materials) I went to a very simple system of C2H4 / C2F4. This is sometimes studied as a very simple system showing charge transfer. I had many issues getting this job to run successfully, before realising that I needed a suitably diffuse basis set for the …
1 min read

1 min read

1 min read

I’ve recently started to use NWCHEM (for access to some modern tasty tasty functionals). I really like the input deck format - much cleaner and more sensible, IMHO than the ‘concise’ header that one uses in Gaussian. However - it looks totally sterile as pure text, so I spent a little while learning how to write my own vim syntax file. While I was at it, I wrote much cruder versions for Gaussian .
1 min read

For my 28th birthday I report on the Bulk Hetero Jelly (BHJ). Unlike conventional Jellies, the BHJ architecture allows for a fully solution processed deposition at lower temperatures and without the vacuum processing necessary for conventional designs. The Molar scale phase separation is achieved implicitly by the kinetically limited dynamics of the slithering snakes and foam strawberries, which are then ‘frozen’ in …
2 min read

A colleague recently asked about assessing the Donor-Acceptor strength of a given polymer by Quantum Chemical calculation. Often the literature just seems to print the pretty electron spin density plots from Gaussian at an arbitrary isovalue, then essentially relying on just making a statement about the 2D area occupation of this plot! I was interested in whether one could find a simple method of putting this on a more quantitative …
2 min read