Figure 1
Tricky problems turn up if anyone uses the human calendar for accurate computations involving earth, the matter of the scatter in the calendar timing of the actual position of the earth in it’s orbit.
A nice little warning “Beware of leap year when predicting climate change, warns researcher” from the Stanford News Service March 2001
This is in insidious problem reaching into many more fields. Post here is because I was reminded the hard way when trying to use conventional math as cross check evidence: I’m expecting screams of dissent when a future blog article is published.
Figure 1 is an XY plot and ideally is a 45 degree straight line. Whilst this is close I would prefer it better. X comes from a model and Y directly from the dataset except for this darn problem. Eye of the needle is something else, no idea what.
I have seen errors in datasets from precisely this cause.
===
Managed to get some sleep last night and kind of wake up today. Nicely today here in England it was less cold, weak sunshine, no rain and less wind. Forecast for tomorrow is… rain and colder.
Couple of days was in a large store looking for what they didn’t have, story of my life. I amused to notice stacked high up almost out of sight were snow shovels and toboggans. Typical shopkeepers, didn’t see snow coming, panicked and overordered at silly prices, proffered at you-must-be-kidding tags, now they have dead stock. EU? Bet they have sell-by dates on them. (saw a plastic wrapped toilet seat being sold cheap, sell by date close!) I tell you, it’s all poo. Maybe someone needs a poke fun web site collecting the daft things which time expire.
Another was a clock but that is funny. Inspector to shopkeeper: This clock is out of time, do you have anything to say?
Psst… “The fire’s tomorrow”, an old solution to old stock.
Posted Tim Channon
Tallbloke's Talkshop 
“Nicely today here in England it was less cold, weak sunshine, no rain and less wind. Forecast for tomorrow is… rain and colder.”
Just like Sydney at the moment. The difference is that it is Winter in Australia and it is Summer in England …
We’re having a traditional winter here in Brisbane, Truthseeker … ones that I remember from years ago. It was my shortlived introduction to AFL, for the life of me the TV always showed games down south in the cold, pouring rain, and muddy paddocks.
Eye of the needle: If the plot is from excel, check to see if the series is formatted to connect data points with smooth curves.
It is close to a loop Terry, XY plot where that is a slight difference in the curves being compared. Been improved slightly since I posted that.
I’m trying to hold the fort here and do a variety of other things.
One of which is novel. This is posted from an OpenBSD box I’m bringing up, a mongrel which will do as an emergency old dog.
I’d been wondering for a while how leap years etc were dealt with. Is there a de-facto standard reference calendar Tim, or is there no single best approach?
Not so far as I know. The point of posting a warmist link was to show one of their kind admitting to errors in their own, from which a little more confidence can be gleaned on who to trust more or less.
The Portal pages on Tallbloke’s has a link to some time calculators. Is a large subject.
Astronomy does the sensible thing and sidesteps, Julian calendar which is purely day based. This is still problematic.
I think there is a widely held belief, naive, usually proportional to hard life experience that experts or loudmouths, or officials… know what they are doing, are right, honest, yada yada.
In reality with time seems to me a particular field does a particular thing. Doesn’t follow it is correct.
When I get around to it I will be posting articles which rip apart a lot of science, climatic, meteorology, where abuse of time is one of the key factors. So far as I can see there is near universal violation of sampled data laws. Yes this matters.
Why for example did the Met Office declare 2011 was very hot when the people knew otherwise?
Answer is very simple and caused by wrong maths. This might have a subtle connection with the first paragraph here in that errors vary with scale. At 10 million years the fact the human calendar spins relative to real time no long matters.
Few take Nyquist and Shannon seriously enough. (I don’t, pragmatic)
Do I get it wrong, make mistakes? Of course.
Tim. If I’m not mistaken, there’s a distinct ‘Irish’ accent in your posts. I hope I’m right in assuming this as, from an anthropological communication POV, it helps me towards a better understanding of what you write. 🙂
Can you convert your data to a ‘sidereal’ type of ‘time signal’ before processing it?
http://en.wikipedia.org/wiki/Sidereal_time
If so, I would expect a less ‘cyclical’ and a more ‘straight line’ in your plot.
Best regards, Ray Dart (AKA, suricat).
Amused. Completely English, Saxon.
If I garble it’s okay to ask. I know I am at times difficult, wish it were not so. (or a choice)
Siderial is fine provided there is no mixing with other time methods, where most climatic datasets using d/m/y.
Also amused, Saxons are Germanic and also Celtic. 🙂
Hm. My Granddad on my Dad’s side was Devonish (Grandma and all of Mum’s side came from Durham), Mum and Dad both came from the Durham region of the UK and I made my first entrance in London UK (I always questioned why my parents spoke differently to everyone else back then). I moved to N. E. Essex at age 13 and was presented with an influx of the Suffolk accent, then ‘fried my brain’ by learning French whilst spending time with the Walloon’s in S. Belgium. Oh, forgot the ‘frying time’ I spent in secondary education learning Russian before moving to N. E. Essex. All told I’m surprised anyone can understand a word I say/write, but I do try. I’m just glad I can use the odd smiley face here and there. 🙂
I tend to respond with the ‘Armed Forces” definition of ‘British English’. It covers a multitude of linguistics and phraseology, but enough of these OT remarks. 🙂
“Siderial is fine provided there is no mixing with other time methods, where most climatic datasets using d/m/y.”
Extra work, I know, but dd/mm/yyyy format can be converted before data manipulation. However, the ‘converter’ choice is limited by the ‘O/S’ and ‘compilers’ in your system.
If you use an ‘Open Source’ O/S you should have ‘Python’ scripting available. If ‘Python’ isn’t currently available, you should be able to find a ‘download’ source for it. MIRIAD looks to be a good app for Python:
http://astro.berkeley.edu/~pkwill/miriad-python/docs/data-util.html
I can’t offer more without knowing the O/S you use and the ‘scripting’ it accepts. You know my email address if this is too sensitive for public disclosure.
Best regards, Ray Dart (AKA, suricat)
Estuary man (leg pull) learning Russian, Waloon…. lets not go there.
For a lot of things I write my own software.
As often as I can get away with it, I use ANSI standard: (yy)yy/mm/dd . It actually sorts properly without code to extract the fields first! Didja know most US dates are mm/dd/yy(yy)?
As for dd/mm/yy, and mm/dd/yy, when was 11/03/10?
In ANSI, that’s 2011 March 10.
________
P.S. its orbit, not it’s orbit. Unless you meant “it is orbit”
his, hers, its;
he’s, she’s, it’s.