Nautilus and Windows shares: wins for the win!

Being one of the very few Linux users in a Windows ecosystem is always a source of little niggles which are annoying but often educational.

Last week, I managed to finally solve a problem I've had for months - reliable mounting of Windows shares on my Fedora workstation.  

Despite having all the required libraries installed and configured (samba etc.), sometimes Nautilus would take 20 minutes (!) to load while trying to connect to my shared drives.  Sometimes it would fail altogether.  And other times it would work perfectly. Sometimes it would fail to start, then open the instant I disconnected my VPN.  Other times it would only work the instant I connected my VPN.  Whether drives would remount after a suspend seemed to depend on local meteorological conditions and my horoscope.  The problem was so changeable that I had real problems working out the issue, or even how to start researching it.

Anyway: once the problem was identified as being netbios, the final solution came from the following blog:

http://charlieharvey.org.uk/page/slow_nautilus_browse_with_netbios

All that was required was editing /etc/nsswitch.conf to change the line

hosts: files mdns4_minimal [NOTFOUND=return] dns mdns4

to

hosts: files wins mdns4_minimal [NOTFOUND=return] dns mdns4

i.e. adding wins support high enough up the priority list.

Looking around support forums this seems like a common problem, but the solution is not so well-known.  Hopefully another blogpost will help spread the news.

 

News worthy of a post

After a couple of months of neglect, I thought I would mark the exciting news that the first ever malaria vaccine has been given the green light by European regulators with a blog post.

I was particularly happy to see the following article on the BBC News from GAVI and the Global Fund, which does a very good job of explaining the complexities that lie behind decision-making in public health, and why we need trials, simulations and statistical analysis to help make those decisions the best-informed we can.  

Fedora and WD My Book hard drives

If you're running Fedora <= 20 and having trouble with mounting a WD My Book hard drive via USB 3.0, take a look at this bug posting:

https://bugzilla.redhat.com/show_bug.cgi?id=1091513

Following the instructions there to remove the timeout on running ata_id fixed the problem for me - i.e. just comment out line #35 of /usr/lib/udev/rules.d/60-persistent-storage.rules

Looking around this seems to be quite a common issue, but the fix doesn't appear to be well known or properly implemented yet.  It's such an obscure but simple fix I thought it was worth sharing.

Jonas Salk and the development of polio vaccine

I really enjoyed this piece in The Guardian last week about the recent 100th anniversary of the birth of Jonas Salk, who developed the most used polio vaccine:

Jonas Salk Google doodle: a good reminder of the power of vaccines

The importance of this vaccine can't be overstated - a devastating, incurable illness with long-lasting effects on thousands of sufferers and their families has been eradicated from almost every country on the globe in just a couple of generations.  It's truly one of the great achievements of public health research.

The article gives an idea of the scale of the undertaking, and the combination of different steps forward from dozens of scientist working in many areas, including e.g. cell line development.  And then I discovered I had yet another reason to recommend it - I noticed it's written by an old school friend of mine, Pete Etchells, who is now a lecturer in Biological Psychology at Bath Spa university and writing regularly for the Guardian.    

Ebola and malaria

There's a very good article on the BBC today about the effect the current Ebola outbreak is having on the fight against malaria in Guinea, Liberia and Sierra Leone.  It's a very unfortunate combination of two problems - the fact that limited resources (particularly medical staff) are now being severely stretched, and that the similarity of symptoms in the early stages of both diseases is leading to reduced treatment seeking.

“These countries have previously been really hit by malaria. But five years ago, it was even worse - the deaths were double.

We all agree that no child should die from malaria, because we have the tools to prevent and treat it.

But now, understandably, all the health workers’ attention is on Ebola.

We used to see hospital beds with three children in them at a time, because there was not enough space.

Now those paediatric wards are becoming ghost areas, because of the lack of manpower there. So we don’t know who has malaria, and who is dying from it. Even if the situation is at the same level as last year, that was still very bad in those countries. We’re really concerned that Ebola will cause a setback to the efforts on malaria.

And there’s a lack of trust and confidence in health workers. There’s still a feeling it’s them who are bringing the virus to people.”
— Dr Fatoumata Nafo-Traoré, Roll Back Malaria, speaking to the BBC

The effect of Ebola on malaria transmission, and of concomitant malaria on Ebola survival, are certainly areas with many open questions.

 

 

Correlation, causation and association studies

Correlation analysis is a quick-and-easy, first-line technique for data exploration in large data sets - but when searching for novel causal relationships, it's only the initial hypothesis generator, before moving to more sophisticated statistical or domain-specific analyses, or targeted collection of more data.

If you ever need a quick and easy demonstration why, have a look at Tyler Vigen's excellent website, Spurious Correlations.  In the last nine years there is a correlation of 0.97 between number of people who died by becoming tangled in their bedsheets in the US, and the total revenue generated by US skiing facilities.

Another example is the negative correlation of -0.93 between honey bees and convictions for cannabis possession:

The last word goes to the excellent xkcd:



The rise of the British urban mosquito

There's a paper in PLOS One detailing an experiment with water butts where it was seen that mosquito breeding grounds in urban environments show higher populations, but lower species diversity.  Interestingly, A. plumbeus (German wikipedia, as there is no English entry yet) seems increasingly to breed in this environment, which has a potential impact on public health as a carrier of West Nile virus.

“Mosquitoes in urban containers were less species-rich but present in significantly higher densities (100.4±21.3) per container than those in rural containers (77.7±15.1). Urban containers were dominated by Culex pipiens (a potential vector of West Nile Virus [WNV]) and appear to be increasingly exploited by Anopheles plumbeus (a human-biting potential WNV and malaria vector).”
— British Container Breeding Mosquitoes: The Impact of Urbanisation and Climate Change on Community Composition and Phenology

The BBC have a news article on the story here.

Malaria Vaccine Initiative in the Guardian

The Guardian newspaper (UK) has run a nice feature which give an introductory overview about the work of the Malaria Vaccine Initiative, for whom I am currently doing some mathematical modelling work.

Malaria vaccines: from impossible to probable

“This week we celebrate the progress being made on both fronts: to expand the reach of existing tools against malaria and to develop and make available effective vaccines. It’s also a time to look forward to what could be—a future where parents no longer lose their children to this preventable disease”
— Ashley Birkett, PATH

Health Atlas of England and Wales

Today there were quite a few news stories about a new Health Atlas of England and Wales produced by the Small Area Health Statistics Unit at Imperial College.

It shows on a very fine resolution the relative risk for a number of important diseases and environmental factors, adjusted for age, sex, deprivation and small numbers.  Some of the patterns have relatively simple explanations (NO2 concentrations high near cities, attributable to traffic and industry), but others are both fascinating and hard to explain.

Relative risk map for malignant melanoma (Environmental Health Atlas for England and Wales under CC license)

Relative risk map for malignant melanoma (Environmental Health Atlas for England and Wales under CC license)

 For example, for malignant melanoma there is a clear high relative risk in Devon and Cornwall which could be partially explained by the greater hours of sunlight it receives - but why is the south east not similarly affected despite its similar sun exposure?  What is the reason for the cluster of high relative risk in Cumbria?

This looks like a great resource for generating hypotheses and testing new analysis methods in spatial statistics. 


New site

Welcome to my new site! 

As you can see, it's very much a work in progress.  Over the next few weeks I will flesh out the research pages with some more detail and links to papers.

I am planning to use this blog mainly for short posts on papers I've found interesting and programming tips and tricks that I've found useful.  But let's see how it evolves...