Is "activity" define in page clicks, posted solutions, posted correct solutions or the sum of problems and solutions? Is it the activity per hour, day or week? Is it the current activity or the integral?
Please add meaningful labels.

Its an awesome book. The other two series that parallel and follow on from it are great btw... I've read it a half dozen times now. Did you know Orson is working on a movie version?

Which one is which? Could you please add a legend?

lol pour la saison 2011-2012 ton code ne marchera pas

My immediate reaction to this is not "Wow look at US debt!" - it is the fine detail of the US data. Although the profiles for the UK and Canada look "flat" is this just because the US reports with more fidelity? It looks like the UK data is fixed to 0.25% increments.

In my experience, the USA is way ahead of the UK when it comes to making data public. In Europe, only the Scandinavian nations rival the US in this area. So take that as a positive at least, MATLAB users of the USA!

Things are getting crazy now.

I've been disappointed with the bacon-related-content of the photos tagged bacon from the public feed. I'm switching to the BACON Pool to see if they're more delicious.

What a bright idea! Good job!

For some reason, the timing of the data collection is causing problems with the results. Its making it look like the highs and lows are getting swapped, but in fact, they are not.

Well, not really useful unless there is a way to get Trendy to update it at 2-3pm instead of 12:15 am. But I tried.

Any idea what is the increase peak of mid-March 2012?

Thank you!

I could help if you had some specific questions. Sorry, i can't do that for you..

I know I'm asking too much, but could you help me in writing these codes?

for year = 2012:1:+Inf
nb_PSG_points = nb_Olympique_Lyonnais_points -3;
end

Of course it is

it is simple to write such codes in matlab?

Mo' playas, mo' problems

That is truly fantastic.

I can has comment?

No, it doesn't. It just parses a number from whatever url you provide.

If you want to filter by months, you have to provide the according url.

this is not that clear. But this counts only the numbers?, I mean, it doesn't change the month automatically, right?

You extract it easily: there is a results count on the web page you can retrieve (just use the urlfilter on it). You won't be able to import the historical data into Trendy, however. Only the new data that comes in.

Legend added! Now no one else will have to struggle with figuring out which curve refers to Hawaii.

Do I get a prize for guessing that Blue is Anchorage?

When you search something on Google news, you can see the number of articles found. E.g. seraching for "tzunami disaster", Google displays 3450 articles. I need to collect it month by month from 1999 to today. Is it possible with matlab?

It's one of my favorite books ever, glad you're enjoying it!

I am trying to figure out how to query correctly and which "questions" to ask.

Could we see:

(Answered + (doit4me tags && not answered)) / total?

It's real! Follow the links back to the trends and you'll see the data comes from SpaceWeather.com. Here is their solar wind explanation: http://spaceweather.com/glossary/solarwinddata.html

I have one quest on this plot.. Is this the data real or just taken randomly?

Thanks Ned!

This is fantastic... great idea!

What is the number of articles?

Rather than using negative numbers, you might want to set the 'YDirection' property on the axis to 'reverse' like so:

set(gca,'YDir','reverse')

It's been a tough couple weeks. Hopefully better days are ahead.

You might want to normalize by the time between samples. If your time sampling is irregular (you miss a day or two, for example), your diff will be thrown off.

Very cool use of URLFILTER. I had to look at the source for the page you were scraping, but now I see why you had to get 12 numbers. Thanks for putting the comment in the code.

I just bought this thanks to this graph :-) It's fantastic!

I would suggest an improvement to this graph: The data points for a particular club should come only from the days that the club has played a game (as opposed to now where it is sourced from when anybody plays a game). This will make it possible to tell the difference between a loss (0 points) and not playing at all that day (0 points).

It would be awesome if home and away games could be distinguished.

Not sure if the dataset would allow these changes though.

Cheers! This was a fun graph!

Hi, do you know how I can extract the number of articles found on Google search?

I'm afraid it won't look quite so authentic when the Times gets more data points. I've printed off a screenshot to hang next to my desk.

Well done sir. Rule Britannia!

(A patriotic idea)

Hmmm, I see your point. I'm quite keen to distinguish zero though. I have an idea.

Would you consider using a different colormap? Something that makes the gradient look monotonic (like the grayscale)?

Just curious, how does GDP per person compare between the two countries?

Down from the heady 100 mark, but still trending up!

oh no! it's slowing down!

This plot got mentioned in The Independent http://www.independent.co.uk/news/people/diary/diary-from-armenia-to-newcastle-and-back-again-6612085.html (see under "Calm down, dears")