Confusing cycle stats - London centric
anonymousblackfg
Posts: 2,029
After some pestering (again) the tfl cycle flow stats are now updated:
http://data.london.gov.uk/datastore/pac ... ad-network
Apart from the usual confusion on why TFL break the year into 13 periods (logical thinking like this may be why they believe cyclists love being under the axels of tipper trucks).
Compared to the same periods last year, we're up 28.3% and 18.5% but when adjusted for 'seaonality' these barely register over a percent?
Regardless, a good push in the right direction.
http://data.london.gov.uk/datastore/pac ... ad-network
Apart from the usual confusion on why TFL break the year into 13 periods (logical thinking like this may be why they believe cyclists love being under the axels of tipper trucks).
Compared to the same periods last year, we're up 28.3% and 18.5% but when adjusted for 'seaonality' these barely register over a percent?
Regardless, a good push in the right direction.
If I know you, and I like you, you can borrow my bike box for £30 a week. PM for details.
0
Comments
-
anonymousblackfg wrote:Apart from the usual confusion on why TFL break the year into 13 periods (logical thinking like this may be why they believe cyclists love being under the axels of tipper trucks).
Because 13 periods means equal 4 week data sets so you are always comparing 20 working days and 8 weekend days. Using calendar months just puts unnecessary noise into the data.Faster than a tent.......0 -
Rolf F wrote:Because 13 periods means equal 4 week data sets so you are always comparing 20 working days and 8 weekend days. Using calendar months just puts unnecessary noise into the data.
Apart from the occasional 53 week year? I assume one of the periods consumes the extra week in those cases.0 -
53 week years? When did that last happen?0
-
Agent57 wrote:Rolf F wrote:Because 13 periods means equal 4 week data sets so you are always comparing 20 working days and 8 weekend days. Using calendar months just puts unnecessary noise into the data.
Apart from the occasional 53 week year? I assume one of the periods consumes the extra week in those cases.
If we have 53 week years then I'm now almost a year younger than I thought I was! It's only one extra day but all that does is mean it takes an extra day to get to the end of the calendar year. It's only an issue if you want to summarise the data to fit the calendar year exactly. You probably could just leave the extra day out or you could just factor it as 21/20 working days or 9/8 weekend days over the final four weeks in December.Faster than a tent.......0 -
365=52x7+1 some will need to be counted as 53 otherwise you'll get a drift so that the (4 week) months are not aligned with a year, takes a while though (2 weeks every 11 years with leap years).Currently riding a Whyte T130C, X0 drivetrain, Magura Trail brakes converted to mixed wheel size (homebuilt wheels) with 140mm Fox 34 Rhythm and RP23 suspension. 12.2Kg.0