I have been making quite a few plots recently for a paper which should hit the arXiv very soon. The Python plotting package Matplotlib has been indispensable in this regard, especially as I took the effort of creating a script which creates all the plots. This meant that redoing all the graphs for new results or with changed sizes etc. was a simple as rerunning the script.

Quite a few of the plots use log axes and while Matplotlib performs admirably there was one problem I had with certain plots. By default, the log plots only show tick labels for each order of magnitude. Tick labels are the numbers on the x or y axis telling you the corresponding numerical value, and when the figure is zoomed in it is possible to lose the major tick label at say 10^-4 because you only want to plot values from 0.3×10^-4 and 0.5×10^-4. Obviously this removes all sense of scale from the plot. A very mediocre solution is to just zoom out until a major tick label is back in the plot but this is obviously unsatisfactory.

I looked through the Matplotlib documentation, which has very detailed information about the API and has a lot of examples, but unfortunately didn’t address this exact point. After a bit of searching I found a useful conversation on the users mailing list which got me close but didn’t use the LaTeX labels which are really essential for publication quality graphs (in my opinion anyway!). The tick labels documentation along with the major-minor ticks example led me to the Formatter classes, especially LogFormatter and LogFormatterMathtext. This looked like the right answer but unfortunately LogFormatterMathtext writes the minor tick labels in a very unusual way. Instead of 0.3×10^-4 it only writes an exponent, so 10^-4.52.

I finally settled on extending the pyplot.LogFormatter class which controls the text for the tick labels. My subclass is as follows:

import re
import pylab

class LogFormatterTeXExponent(pylab.LogFormatter, object):
    """Extends pylab.LogFormatter to use
    tex notation for tick labels."""

    def __init__(self, *args, **kwargs):
        super(LogFormatterTeXExponent,
              self).__init__(*args, **kwargs)

    def __call__(self, *args, **kwargs):
        """Wrap call to parent class with
        change to tex notation."""
        label = super(LogFormatterTeXExponent,
                      self).__call__(*args, **kwargs)
        label = re.sub(r'e(\S)0?(\d+)',
                       r'\\times 10^{\1\2}',
                       str(label))
        label = "$" + label + "$"
        return label

It is provided as is, but there shouldn’t be too much wrong with it. One odd thing is that the LogFormatter class is an old style class, so I inherited from object to make it my subclass a new style class. This might be dangerous and cause some unexpected problems.

To use the class you can do something like the following:

import pylab
import numpy as np

fig = pylab.figure()
pylab.semilogy(np.logspace(-6,-5))
ax = fig.gca()
ax.yaxis.set_minor_formatter(
    LogFormatterTeXExponent(base=10,
     labelOnlyBase=False))
pylab.draw()

Below are three different figures showing the current default situation, the result of using LogFormatterMathtext and the result of the new class. I hope this will be of use to someone who has been struggling with this problem as I have.

As I mentioned, this came up because of a paper that is very nearly completed and should be available soon. Along with the paper we should have the long promised release of the code I have been working on which solves cosmological perturbation equations during inflation. More on that soon.

The website for the current beta phase of the project is http://inspirebeta.net.

It is unclear to me whether “Beta” is part of the site name, as suggested by the URL and the text on the INSPIRE page, or whether this is just the beta phase of the INSPIRE project as the SPIRES homepage seems to imply. It would certainly be an odd decision to use a different URL for the beta phase and force everyone to change bookmarks, references in blogs, literature etc., once the beta phase is over.

These are quick first impressions because I haven’t had much time to use the new service. First off it is fast. Very fast. So fast that when searching for my name INSPIRE claims that the “Search took 0.00 seconds”. It feels almost instantaneous. This might be because of a light load before the hordes using SPIRES are switched over. It is certainly an improvement on the interminable and often futile stretches of time needed with the SPIRES engine.

A quick glance at the new search features shows that “Google style” text searching is now accepted, so no longer will the archaic “fin a Name” syntax be necessary. However if you are wedded to the old SPIRES way of operating then this is also supported. There is good support for eprint search using the “arxiv:” operator, which lead me to the interesting discovery that the arxiv.org site parses addresses of the form “arxiv.org/abs/arXiv:XXXX.XXXX”. In other words, including “arXiv:” after the final slash still resolves to the correct paper even though the canonical way is “arxiv.org/abs/XXXX.XXXX”. (It also works with lower case “arxiv”.) This is quite a handy feature once you realise that the Bibtex output from INSPIRE now populates the eprint field with “arXiv:XXXX.XXXX” instead of just “XXXX.XXXX” as SPIRES does. There is also a regular expressions search mode which should be very useful for power users. Limited search inside articles seems to be supported as well.

The article page is vastly different to the old SPIRES listing. A new tabbed panel takes centre stage, offering “Information”, “Citations”, “References” and most intriguingly “Plots”. This last tab displays all the graphics from the arXiv version of the paper. (I think this is the case purely because the text on the homepage beside Eprint Number says “Note the plots”. And the copyright difficulties from harvesting journal figures.) The screengrabs in the gallery below are all of this paper which is linked to from the INSPIRE front page and will probably gain at least a few more citations as a result of the exposure.

It will take time to see whether this new update to SPIRES remains as fast as this early test promises. The original news stories also promised some social connectivity which seems to be currently lacking when compared to the arXiv’s array of social links. Overall this is a positive new beginning for HEP article and citation searching.

Update Dec 2011: The full Inspire website has been operational for a while now at its new address of http://inspirehep.net.

As an Android user for a while now, the options were previously limited to navigating to the arxiv website which doesn’t really scale well onto a phone screen, and manually searching for the right paper, or scanning the new list.
For a while there has been an iPhone app called arXivew which listed the latest papers and allowed you to easily download and view the one you wanted.

Now, there is an Android app to compete. ArXiv Droid, released in June by Jack Deslippe, performs admirably, listing all the new papers in each category, including sub-categories like astro-ph.CO for cosmology. You can select categories as favourites for easy access and easily select and view the abstracts of papers. Papers are viewed using the builtin pdf viewer so your mileage may vary depending on whether you mind scrolling around the page. Handily the app prominently displays the size of the pdf before it is downloaded, in case data limits are tight.
Paper titles can be shared via the usual methods, but there doesn’t currently seem to be a way to select papers as favourites, or to see a log of the last abstracts viewed (there is a list of pdfs viewed). However, having seen the progress made in just the last few months, I would look forward to seeing something like this in the future.

With the exception of the problems of viewing A4 sized pdfs on such a small screen, this app does a great job of getting you to the papers you want fast when you are away from your desk. As the proliferation of tablets, e-readers and mobile units continues, I think the arXiv might have to consider some kind of additional format suitable for use on these devices. A quick fix might be to just compile the LaTeX with the geometry of the page changed to A5 or smaller. This would obviously still not allow the reflowing of text which is a key feature of e-readers, and would probably mess up a lot of formatting rich papers.

Until this problem is solved arXiv Droid is an admirable app which helps you stay on top of the deluge of research now flooding the arXiv every day. Get it in the Android Market from this link or by scanning the QR code.

I use the Beamer package for LaTeX which is a great way to include mathematics in your slides, and is pretty straightforward to use if you are proficient with LaTeX.
The default settings in Beamer are quite pretty but after a few days of a physics conference they can become quite repetitive. I like to make my slides at least somewhat different to all the others out there, and try to use my customisations to keep the attention of my audience.

One main change I think is useful with beamer is to make the equations use a serif font like a normal paper would, and not the sans-serif font used by default. This is achieved by having the following command
before \begin{document}:

\usefonttheme[onlymath]{serif}

To get rid of the navigation symbols at the bottom right of each slide I use

\setbeamertemplate{navigation symbols}{}

and the template I use is given by

\usetheme{Frankfurt}
\usecolortheme{rose}
\usecolortheme{seahorse}

I don’t like all the clutter that’s normally at the top of each slide
(contents, title etc) so I use the “plain” option for each frame:

\begin{frame}[plain]
...
\end{frame}

To do the black background with white text is slightly tricky but the
template below should work. Just be aware that the curly brackets
outside all the other commands are required to limit the change to
just one frame.

{
\setbeamercolor{normal text}{bg=black}
\setbeamercolor{whitetext}{fg=white}
\begin{frame}[plain]{}
\begin{center}
{\usebeamercolor[fg]{whitetext}

INSERT TEXT HERE

}
 \end{center}

\end{frame}
}

These are pretty simple changes but used judiciously they can have a striking effect. I think the most useful is the change of maths text to be serif, in line with the standard used in academic print. The sans-serif maths font just looks a little odd in comparison.

After last week’s Science Online London 2009 conference which I attended, I have been thinking about how to get fellow cosmologists to start interacting online. I am not sure whether anyone else will use it but I have started using the hashtag #cosmo09 on twitter and have created a FriendFeed room for the conference. There might not be much activity, but if people do want to use these tools, at least they will have somewhere to start.

So, it may come as something of a shock to learn that I have just spent some such time away from my work, honing those tenuous skills, and have come back re-energized and full of enthusiasm. I spent three (and a half) days last week in sunny Bournemouth, at a UK GradSchool, organised by the UK Grad team (soon to be known as Vitae). This consisted of team building exercises, project management tasks, interview workshop and an outdoor component to bring it all together.

I hope I don’t give too much away, but the main thrust of the week was solving different problems and facing different scenarios in small groups of about 6 or 7 PhD students. Tutors, with a wide range of career and personal experience, helped us learn from each exercise and guided us through the emotional experience of a new team being formed. It’s hard to describe what working with 5 other PhD students from wildly varying areas felt like, but it was definitely intense. By the end of the week, people had gone through more with the others in the group than perhaps they ever had with those they work with every day. In particular the opportunity to give and receive individual and honest feedback on how we affected those around us was surprisingly powerful.

Interview skills were explored in a task designed to test students as both interviewees and panel members. Sitting on the other side of the desk really highlighted how much of the process is about the applicant selling themselves. It was hard enough to distinguish three candidates answers from each other after a long morning, so making an impression is clearly important.

Overall, my experience of GradSchool has completely brushed aside any reservations I had about it taking up valuable time. I may not measure last week in terms of words written or papers read, but the skills learned (and hopefully friendships made) will make the coming year much more manageable.

To learn more about the GradSchool program visit the introductory page at UK Grad, but be warned that word has spread and courses are booked out months in advance!

The new academic year has brought with it some new tools from one of my favourite web resources Cosmocoffee. As you can read in this forum post, there are three new additions to the service. Firstly new search options are available which allow you to use the search page on Cosmocoffee to search the arXiv, ADS and Google Scholar. I don’t know how useful this might be, as I tend to use the integrated search bar in Firefox to directly search SPIRES and the arXiv.

The main update however is the addition of a bookmarking system to the arXiv listings. While not as fully featured as either Citeulike or Connotea, this is a very intuitive system and can be easily integrated into your workflow if you already use Cosmocoffee to access new arXiv papers.

The final tool is a complimentary function of the bookmarking system, allowing multiple users to share lists of bookmarks in a “Journal Club” system. There is a rudimentary management system, with the ability to add users and other managers, and move papers into “old” and “ignored” categories. There is also an anonymous list of all the papers that have been bookmarked so far, which provides an interesting insight into the reading habits of Cosmocoffee users.

To use the bookmarking system you will need to register at the Cosmocoffee site. Since last year registration has been restricted to people affiliated with academic institutions.

Paper hunting:

• The arXiv – You aren’t going to get far in physics without having heard of the arXiv, but it deserves a mention for getting rid of trips to the library.
• SPIRES – For article searches in particle physics/astrophysics, SPIRES is the last word. The search syntax is a little more involved than a Google search (“find a authorname and j journalname” etc.) but the information available for each article is worth the effort. Particularly valuable is the BibTeX entry for each article, and links to both arXiv preprints and e-journals.
• Google Scholar – Not specific to physics, this is Google’s take on academic search. It catalogues the main journal indexes, and has some useful features like the “Related Articles” search, which does a good job of finding other articles with similar subjects.

Categorizing papers:

In the old days, researchers had piles of papers on their desks, under their desks, and generally all over the place. But if required they could pick a required paper out of this filing disaster quite easily with a good memory and a little luck. Today most of the papers you read might remain out there on the network with only a select few qualifying for ink and paper. How do you remember which papers you’ve read, where they are and what you thought of them?

• Citeulike – This site allows you to add papers to a personal list, add tags to describe the papers, and provides automatic links to the electronic versions. When looking at an abstract on the arxiv for example, you simply click a bookmarklet and the title, journal etc are automatically added. You are also able to rate papers, and export a BibTeX list of all your papers.
• Connotea – This is the Nature Publishing Company’s effort at an online reference manager. As with Citeulike a bookmarklet is used to add papers to your collection. The export functions also allow you to also use a desktop based reference manager such as Endnote.
• The Academic Reader -
  This is a very new site that hopes to offer a portal to many different
  sources of scientific articles. Unlike Citeulike or Connotea you read
  the abstracts on the site itself and don’t have to deal with
  bookmarklets. It also provides a “Library” where you can store
  references to papers you have read.
• Del.icio.us – This is not a science specific tool, but rather a handy social bookmarking system. You give webpages tags, can view other users’ saved items (while also being able to hide selected items) and use “Live bookmarks” of the RSS feed of tags to access your bookmarks in your browser. Now with the new Firefox extension, this functionality is integrated seamlessly into the browsing experience. A lot of people tag abstract pages so they can go back to get the file any time, many using the arXiv tag.

Community:

A large part of doing research, or so I’m told, is becoming part of the research community, communicating with your peers about your work and networking to form possibly collaborative relationships.

• CosmoCoffee – For cosmologists, this site provides a forum for discussions of recent papers and general queries. There is some integration with the arXiv, allowing search and BibTeX retrieval (handy as this is not provided by the arXiv itself) but also keyword based filtering of the latest papers. This allows you to concentrate on papers relevant to your work, especially from large sections like astro-ph, which is getting so large it’s easy lose your way.
• Nature Network London – This is the Nature Publishing Group’s attempt at a social network for scientists. It only started recently, so there is not that much activity yet, but it does seem to have a few people writing and networking.
  [Update] PZ Myers points to a good discussion of the benefits of social networking for scientists and the Nature Network in particular.

So not an exhaustive list, but hopefully there are a few useful resources there. As I said above, if there are any other sites you would recommend please let me know in the comments section.