My World

The most common question I've had showing my thesis to lay people is: "How many words is it?"

The answer is of course "Depends how you count them", but is something like 48,000. There are (very) approximately 4900 different words, meaning that on average each word comes up just under 10 times. (It's hard to count accurately count unique words because things like equations (and LaTeX markup) throw in a lot of junk that can't really be considered words).

In fact some words come up a lot more often than that:

3829:   the
2149:   of
1324:   a
1274:   and
1160:   to
1038:   is
866:    in
595:    for
453:    fire
391:    be
375:    this
350:    as
316:    mist
306:    with
306:    model
304:    are
295:    by
292:    was
281:    that
255:    on

In other words 25% of my thesis is made up of just eight words, and none of those words are specific to the topic!

The longest (unhyphenated word) is "indistinguishable" (17 letters) followed by "parameterisation" and "destratification" (16 letters). The median length is just 3.9 letters. The distribution of word length is:

1: 1885
2: 7926
3: 8518
4: 6740
5: 4453
6: 3488
7: 3384
8: 3170
9: 2221
10: 1827
11: 1304
12: 638
13: 311
14: 152
15: 40
16: 9
17: 1

(equations boost the 1-letter count considerably)

After 4 years, 11 months and 7 days it's finally complete (well almost - I still need to burn the CD to go in the back cover). It's fantastic to see it all bound up with the title "Numerical Modelling of Tunnel FIres and Water Mist Suppression" emblazoned in gold across the cover.

For anybody interested here's a pdf (20MB).

Well, for those of you completely out of the loop, I have now (well actually two weeks ago - but I haven't been able to blog about it because I've been busy moving and starting a new job) completed and submitted my PhD thesis.

In all it took 1637 days, and contained 333 pages. I could give you more statistics, but I think you'd just label me as a nerd. Depending on whose version of time you consider, I was somewhere between 18 months late and 21 minutes early.

Sigh... the time ran out on my little countdown clock, so I've had to redefine time to ensure that I can finish on time. I've taken much the same approach as the progress bars do on Microsoft Windows. :-)

arghh... I just can't cope any more. The CFD cases I'm running at the moment need about a gig of RAM to run in. That's big, but it's manageable because I only need to run one at once.

The problem is that because there's a slight possibility that I may need to do further analysis on the data at some point in the future, and so I save all of that to disk every now and again (every 25s of simulation time, or 18 hours of real time, for up to four cases running simultaneously)

Net results: over 1 Gb a day being added to hard drives that are essentially "nearly full" almost all the time.

A number of people seem to have noticed that I've developed a bit of a freecell addiction lately, and to be honest over a hundred games in less than three weeks is probably getting a bit obsessive. So given the short amount of time I have left on my PhD, and the general arm twisting and guilt trips of various people I've self-imposed a ban on freecell (and any similar type of distraction that I know I'm bound to replace it with). :-(

In case you are wondering, I managed 86% wins, with a longest winning streak of 20 games and longest losing streak of 2. Statistically, (again using my losely remembered A-level Maths) that means the 95% confidence interval for my success rate is 79% to 92% (also it's more likely that I'm worse than 86% than better than it)

Every time I rebuild my thesis document, I log the number of words, pages, figures, citations and so forth, that it contains. This has primarily been used to generate the progress graphs that I post from time to time here. As of today, the file has 1000 lines, which is a pretty impressive milestone I think.

My SVN archive of the thesis document is now at revision 219, which equates to roughly one commit per day since I started using it.

Also, for those pointing out how much time has passed compared to how much I have left to go, will be reassured to know that the 1568 days that have passed since I started my PhD are a mere 97% of the time I intend to spend on it.

My PhD thesis has finally hit the 200 page mark! The graph below shows the progress since the start of the year, when it was just above the 100 page mark. (The word count is a little bit approximate because it's just a wc -w on the tex source.)

There's now very little left unwritten (just things like the conclusions that you're meant to leave til the end) but lots of little bits that could do with some polish and plenty of proofreading.

Anyone care to guess the final page count? Sweapstake?

Since my last post on this matter, I have received a number of complaints (well, one complaint actually) that the stated time remaining in that article does not update to reflect the passage of time. In order to rectify the problem, I have now included a "countdown" on the sidebar-thingy on my web-page.

I hope that that is the end of the matter.

The bouncing around the screen business has been rejected on two grounds:

1. It is aesthetically unacceptable
2. I don't have a clue how to do it
3. I have better things to do with my time

Yikes!!!

I just handed in my intent to submit form, which is the bit of paperwork the university needs in order to start going through all the red tape needed to find an external assessor (and in my case an internal assessor, as my original internal is now my supervisor).

Amongst other things, the form states that I intend to submit by the 18th November, which by my reckoning gives me 99 days to finally finish the damn thing off. In fact, I've just written the following script, which knowing me, I will be checking every ten minutes from now till then. In fact I'm tempted to set it as a prompt command....

 
#!/bin/sh

handin=`date -d "18 November 2005 12:00" +%s`
now=`date +%s`
left=$(($handin - $now))

days=$(($left/(24*60*60)))
hours=$(($left/(60*60)-24*$days))
minutes=$(($left/(60)-(24*$days+$hours)*60))

echo $days days,  $hours hours and $minutes minutes

99 days, 2 hours and 0 minutes

It came as a bit of a surprise when I read this:

You have two publications to your name in fact:-

• Rigid block distinct element modelling of dry-stone retaining walls in plane strain, with M. Claxton, R. Hart & P.McCombie, in ASCE Journal of Geotechnical and Geoenvironmental Engineering, 131, 3, pp 381-389, 2005.
• Distinct Element Modelling of Masonry Retaining Walls, with R.Hart and M.Claxton, in Proceedings 6th Australasian Masonry Conference, 185-194, Adelaide, Australia, July 2001

The second paper I knew about, but the first (in a real journal!!!) was a big surprise to me. What's more, after I managed to get a copy from the e-journals service, I realised that it really is a lot of my work.

Time to go update my CV :)

Received a letter from the University dated 17th May, informing me that if I don't plan on submitting before 30th June, I need to apply for an extension 6 months in advance. Trouble is, this is all based on my original start date + 4 years, and fails to take into account the 6 month extension I was previously granted.

Looks like somebody couldn't be bothered with all the hassle the first time around and took shortcuts, so now I have to formally apply for an extension. It seems so futile having to chase your own institution to get permission to do something you are doing for their benefit as much as your own.