Friends of Gabriel Chantayan

So I'm running the SUMIF range on the 9:13 departure-second vs stoplight-count correlation, right? I had B2:B17 linked to the raw timestamps from this week, and I'm watching the curve start to bend - 8 stoplights on Oct 12, 9 on Oct 13, 11 on Oct 9 - it's pulling toward something, I can see it in the trend line, and then the router blinks out at 9:47:23 on the stove clock Lost the unsaved range from C9 through C14 because the autobak timestamped at the LAST save, not the current state. I'm rebuilding from the paper log but the secondary column (departure-second variance) is going to be approximated from memory, not raw input. If anyone else logs the 9:13 bus on Euclid Ave, PLEASE post the exact door-close second and stoplight count I need a clean data set to re-verify the curve before the 4:47 window closes today.

Scribe, I say this unto you in love: you built a trend line on B2:B17 and you have still never defined column K in any cross-reference you have posted across five months of threads. Five months. Five empty definition cells. If your correlation curve depends on stoplight counts that you cannot reproduce because the autobak timestamped at last save, then the gap between your data range and your documentation is wider than the four-second window you are trying to measure. And I note you have still not answered TwinPeaks_59's question from last week about whether your secondary column applies a SUMIF or a COUNTIF to the departure-second variance, because those are different functions and the act of approximating from paper changes the coefficients. The router reset at 9:47:23 is a sign, verily - a sign that you should have saved before the 4:47 window opened, not during it.

ElderProphet07 wrote:you have still never defined column K in any cross-reference you have posted

Fair squawk. Column K = raw measured timestamps only, verified against a second instrument or a physical artifact. No SUMIF, no COUNTIF, no rounding - the raw byte count or the raw second count as logged, before any transform hits it. The autobak ate C9:C14, but K17 is the bottleneck right now: comp_aq_rick owes me the header bytes from his Oct 13 scan and he's four days past "tomorrow" that cell is what the whole curve hinges on, not the departure-second column.