No time for actual test runs, and can't afford to go testing. So I'll ignore it when I don't have enough data. I'm only really interested in sorting the mixture at points that I actually use. And curing the odd misfire that afflicts us at times.

Thanks for your uneducated insight though

Quote from tristancliffe :

Thanks for your uneducated insight though

ok so let me get this straight
youre trying to remap the ecu based on an uneducated guess from absolutely no data at all and arent willing to drive the car at different throttle positions (which would probably take less than half a lap) long enough to get some useful data and im the one acting uneducated?

No, I'm trying to improve the map of the ECU where I have data that is useful, knowing that half a lap would get me next to no data to actually use. Remember, the data I am using is based on 15 minutes of driving - including slow out laps, warming up, quick laps, and a slow drive back to parc ferme.

You are acting uneducated because you clearly know very little about what I'm attempting to do, yet you are claiming (yet again) to be the messiah on the subject.

where did i ever claim to be the messiah when it comes to ecus?
by profession i am however fairly adept at data analysis and able to tell when data is sparse enough to be almost completely meaningless

A 6 year old would know that the information is only valid with enough data. I don't see where I say I believe the entire graph??

Dear god, shotglass, have you turned into a doomy naysaying troll in EVERY thread, or just the ones I've finished up subscribed to?

Quote from tristancliffe :

I don't see where I say I believe the entire graph??

you were pretty excited about the graphs that came out of this
also did you ever bother to calculate the error bars for those graphs? considering how noisy the signal is id expect them to make any averaged surface plot almost entirely meaningless

Pretty excited about managing to create the graphs I wanted is different to belieiving every box in the table at the end...

And no, I haven't calculated the error bars. I don't need to - I can see which data is valid, and I know the areas of interest (hint: it isn't where the data is sparse.

You do realise that I got over 65000 points of data from each session of Mallory?? That's more data samples then some actual scientific thesis are based on...

Quote from tristancliffe :

And no, I haven't calculated the error bars. I don't need to - I can see which data is valid

terribly sorry about not feeling awefully confident about the data analysis capabilities of someone whod use excel for... well anything

Shotglass knows about everything. He has had experience in all subjects. From phone anthennas to data handling.

Quote from Shotglass :

terribly sorry about not feeling awefully confident about the data analysis capabilities of someone whod use excel for... well anything

Terribly sorry for not caring what you think about my hobby or the way I go about it. Must be awful being you.

Tristan, is there a calculable optimal AFR per throttle/rev? If so, it might be interesting - or even more visually useful - to plot the AFR anomaly rather than just the mean data.

That's what I've been doing - since the kind the helpful people in this topic got me on the right path of being able to visualise the data, I've been showing the difference to ideal in a table, which is colour coded. I'd also made it ignore points where there wasn't much data (which Shotglass said I wasn't doing), so that I'm only making changes where there is valid data.

And, of course, I'll also tend to look at each lap in a standard graph of values versus time (or distance) like a normal datalogging plot.

I plan to update the map a bit tonight or tomorrow, and see what it's like at Oulton Park on Saturday... I can't have much bad luck left over, can I?

Quote from tristancliffe :

That's what I've been doing - since the kind the helpful people in this topic got me on the right path of being able to visualise the data, I've been showing the difference to ideal in a table

Oops.. not been paying close attention. Cool!

Quote from tristancliffe :

Must be awful being you.

thanks for your concern but its actually quite ok