computer averge speed algorithm

Andrew Webster <[email hidden]> wrote in message
"]news:[email hidden]...

Quoted message said:

"AndyK" <[email hidden]> wrote in message

news:<[email hidden]>...

Quoted message said:

<cut>

Quoted message said:

The "computer" used probably doesn't have the ability to divide - so

instead

Quoted message said:

Quoted message said:

it's using some sort of mathematical "trick", effectively adding the

current

Quoted message said:

Quoted message said:

speed onto the accumulated average, using a weighting factor. Or

sumfink

Quoted message said:

Quoted message said:

like that :-)

AndyK

My cateye unit certainly does something odd, I can be riding along steadily at below my current
average speed and see the average speed go up.

I'd say that was normal.

Quoted message said:

When average is increasing it often goes up by 0.2 more a minute or so, then down by 0.2 then up
0.1 again.

My Sigma displays current MPH integer only, but measures it in 0.1 increments, as indicated by the
max speed function. The average speed seems to be measured over a relatively small time or distance
increment based on the 0.1 MPH increment, therefore, if I have a steady 9 MPH 'indicated', over an
hour, this may be marked with a + , 'dot' or - sign many time and at any given time (indicating that
9MPH is below, equal to or above my current average speed) . The average speed seems to be
dynamically re-calculated and frequently so.

FWIW, HTH - HAND! ',;~}~

Shaun aRe - IMHO, AFAIR, IIRC etsuckingfetera.

Shaun Rimmer <[email hidden]> wrote in message
"]news:[email hidden]...

Quoted message said:

Andrew Webster <[email hidden]> wrote in message
"]news:[email hidden]...

Quoted message said:

"AndyK" <[email hidden]> wrote in message

news:<[email hidden]>...

Quoted message said:

<cut>

Quoted message said:

The "computer" used probably doesn't have the ability to divide - so

instead

Quoted message said:

Quoted message said:

it's using some sort of mathematical "trick", effectively adding the

current

Quoted message said:

Quoted message said:

speed onto the accumulated average, using a weighting factor. Or

sumfink

Quoted message said:

Quoted message said:

like that :-)

AndyK

My cateye unit certainly does something odd, I can be riding along steadily at below my current
average speed and see the average speed go up.

I'd say that was normal.

Quoted message said:

When average is increasing it often goes up by 0.2 more a minute or so, then down by 0.2 then up
0.1 again.

My Sigma displays current MPH integer only, but measures it in 0.1 increments, as indicated by the
max speed function. The average speed

seems

Quoted message said:

to be measured over a relatively small time or distance increment based on the 0.1 MPH increment,
therefore, if I have a steady 9 MPH 'indicated',

over

Quoted message said:

an hour, this may be marked with a + , 'dot' or - sign many time and at

any

Quoted message said:

given time (indicating that 9MPH is below, equal to or above my current average speed) . The
average speed seems to be dynamically re-calculated

and

Quoted message said:

frequently so.

Uhhhm, I meant to write something else here as well:

If your current average speed is based on you travelling 5 MPH (for instance) for 1/2 and hour, then
travelling at 10 MPH for half an hour, this gives 7.5 mph (if my currently addled brain is thinking
correctly) average over the hour, and the comp should indicate that, at the 1 hour mark. If you then
slow down to 7 MPH (0.5 MPH below the curect average), you are still increasing your trip average
speed for the next half-ish hour hour of riding (someone who's more practised at this would surely
be more precise, but hopefully, I'm not wrong again......).

I'm going for a lie down now.........

Quoted message said:

FWIW, HTH - HAND! ',;~}~

Shaun aRe - IMHO, AFAIR, IIRC etsuckingfetera.

[email hidden] (Andrew Webster) wrote in message
news:<[email hidden]>...

Quoted message said:

My cateye unit certainly does something odd, I can be riding along steadily at below my current
average speed and see the average speed go up.

When average is increasing it often goes up by 0.2 more a minute or so, then down by 0.2 then up
0.1 again.

I wonder if this is what Intel did with all those reject Pentiums that couldn't divide properly.

Andrew Webster

I have two Cateye's (Micro -90 and Mity II -95) and they both do the same thing. I have noticed that
the AVG goes up in increments of 0.1 and then drops about 1%. E.g. if I ride constant 30.0 km/h, it
shows an AVG of 30.0 -> 29.7 -> 29.8 -> 29.9 -> 30.0 -> 29.7 etc. From these values the largest one
is the most accurate. When I press the stop button the unit shows an AVG that is between -1% to + 0%
of the accurate value.

I believe this inaccuracy is due to power and production cost savings. All small microcomputers use
integers and using more accurate integers demands a more expencive prosessor and/or more power,
which is very limited with batteries. Since the trip and clock are saved quite accurately (1 s and
10 m), the rounding (or more likely cut-off) error has to come from the calculation of AVG. On a
small microprosessor dividing and multiplying is very time consuming. A single div or mult operation
might take 100-200 clock cycles while adding takes only 1-4 clock cycles. The prosessor takes "much"
power while active, thus it is put to sleep whenever possible. The power consumption depends on the
fraction of time it has to be active. Thus one has to make all the calculations as simple
(=inaccurate) as possible.

Joakim

Quoted post said:

Originally posted by Shaun Rimmer
Uhhhm, I meant to write something else here as well:

If your current average speed is based on you travelling 5 MPH (for instance) for 1/2 and hour, then
travelling at 10 MPH for half an hour, this gives 7.5 mph (if my currently addled brain is thinking
correctly) average over the hour, and the comp should indicate that, at the 1 hour mark. If you then
slow down to 7 MPH (0.5 MPH below the curect average), you are still increasing your trip average
speed for the next half-ish hour hour of riding (someone who's more practised at this would surely
be more precise, but hopefully, I'm not wrong again......).

I'm going for a lie down now.........

Nope - if your true average speed is 7.5 (which it would be in your example) and then your true current speed drops to 7.0, your true average speed will start dropping immediately. It is always the case that if your true current speed is less than your true average speed, your true average speed will decrease; anything else that shows up on your computer is round off error or sample error or possibly algorithm error.

Have a nice nap.

DiabloScott <[email hidden]> wrote in message "]news:[email hidden]...

Quoted message said:

Shaun Rimmer said:

Uhhhm, I meant to write something else here as well: If your current average speed is based on
you travelling 5 MPH (for instance) for 1/2 and hour, then travelling at 10 MPH for half an

hour,

Quoted message said:

Quoted message said:

this gives 7.5 mph (if my currently addled brain is thinking

correctly)

Quoted message said:

Quoted message said:

average over the hour, and the comp should indicate that, at the 1

hour

Quoted message said:

Quoted message said:

mark. If you then slow down to 7 MPH (0.5 MPH below the curect

average),

Quoted message said:

Quoted message said:

you are still increasing your trip average speed for the next half-ish hour hour of riding
(someone who's more practised at this would surely be more precise, but hopefully, I'm not
wrong again......). I'm going for a lie down now.........

Nope - if your true average speed is 7.5 (which it would be in your example) and then your true
current speed drops to 7.0, your true average speed will start dropping immediately. It is always
the case that if your true current speed is less than your true average speed, your true average
speed will decrease; anything else that shows up on your computer is round off error or sample
error or possibly algorithm error.

Have a nice nap.

Heh.....seems I needed it! ',;~}~

Shaun aRe gotz his excuses - he's a head full of wedding/honeymoon plans at the moment,
hehehehwhooooooo! Ahem, ',;~}~