Thanks, Daniel. Based on my understanding of your explanation, I'll choose to trust the raw numbers and calculate the percentages based on those numbers. Do you concur? BR, Dave
On Mon, Feb 8, 2021 at 11:51 PM Daniel Gruno <humbed...@apache.org> wrote: > A quarter in this scenario is defined as "the 3 months preceding this > day", as projects typically report every 3 months on a somewhat > arbitrary day (3rd wednesday), and some projects are asked to report > again the following month. > > Thus, it's highly likely that the reports were not made _exactly_ 3 > months apart (usually there can be up to a weeks worth of hidden extra > data due to the 3rd wednesday rule), and therefore differ a bit in the > numbers that are used. > > For instance, one report could be filed on March 10th, and the next on > June 17th, leaving the time between 10th and 17th to skew the stats. > > HTH. > > On 08/02/2021 20.27, Dave Barnes wrote: > > On my project (geode.apache.org), some of this quarter's change > percentages are way off, if they're comparing to the previous quarter. > These percentages should be near zero, yes? > > > > Today’s stats: > > - 322 PRs opened on GitHub, past quarter (-13% decrease) > > - 321 PRs closed on GitHub, past quarter (-12% decrease) > > > > Previous quarter: > > - 324 PRs opened on GitHub > > - 325 PRs closed on GitHub > > > > Or am I misunderstanding how the percentages are figured? > > > > > > --------------------------------------------------------------------- > > To unsubscribe, e-mail: dev-unsubscr...@community.apache.org > > For additional commands, e-mail: dev-h...@community.apache.org > > > >
