Path: csiph.com!news.mixmin.net!eternal-september.org!feeder.eternal-september.org!.POSTED!not-for-mail From: Anton Shepelev Newsgroups: sci.stat.math,sci.physics.relativity Subject: Re: statistics in Roberts. Was: RAW vs. raw image format Date: Mon, 20 Mar 2023 14:07:41 +0300 Organization: A noiseless patient Spider Lines: 80 Message-ID: <20230320140741.d36db96c045f38b3bd2ff2c9@g{oogle}mail.com> References: <20230219220058.8d3d14741e18cce1bf19e256@gmail.com> <51151e80-a719-46ef-8095-6535309e7d02n@googlegroups.com> <20230220003936.ca90df6f8848a095271a0cbe@gmail.com> <20230223193132.41882edd1d9110b60e745dac@gmail.moc> <20230225001353.60271597ed5a42bec16e8d54@gmail.moc> <0u3qvhlnu50kk3kg7e7jn6ujnene2fo8jk@4ax.com> <20230319004103.a1d8cad77b443543374dc671@gmail.moc> MIME-Version: 1.0 Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: 7bit Injection-Info: reader01.eternal-september.org; posting-host="72f5a17706e17b32348d2757747e7090"; logging-data="3688796"; mail-complaints-to="abuse@eternal-september.org"; posting-account="U2FsdGVkX183voW8Gny35mjTvD6Gwx+LR13FfzxBeCQ=" Cancel-Lock: sha1:P4g9rmH3NmL6MHuGbo0dAyERvFc= X-Newsreader: Sylpheed 3.7.0 (GTK+ 2.24.30; i686-pc-mingw32) Xref: csiph.com sci.stat.math:10872 sci.physics.relativity:604775 Tom Roberts: > > The Miller data /are/ a time series in a way, with the > > readings as uniform as the rotation of the device. [...] > > Each run can be considered to be a (regular) time series > ONLY if it has no omitted adjustment turns. If regular means gapless. In my pre-processing, I incorporated the adjustments as you did but instead of cutting the adjustment turns out I counted them, producing time series with gaps in readings. Thankfully, there are methods and even software to analyse such series. > From the dozen or so of those that I have seen [#], I can > say that you are HIGHLY unlikely to find any significant > 1/2-turn or 1-turn "signal" in them. You published set contains several runs marked "steady", "very steady", and "perfectly steady" in your transcriptions. They are not, however, literally monotonic, but often predominantly so. > The ones I remember were simply a monotonic drift totaling > less than 1 fringe during 20 turns this is a self- > fulfilling prophecy, because if it drifted much more than > that, he would have made an adjustment. (My memory 17 > years later is hazy.) Of course, yet there are "perfect" Mt. Wilson runs with no adjustments and a considerable (in comparison to others) half-periodic signal, present before any comb-filtering. > When I was at CWRU to give a physics colloquium on this, > Prof. Fickinger and I spent a few hours in the archives > scanning Miller's data sheets for runs with no > adjustments. Our only selection criteria were 20 turns > with no adjustments (some runs have fewer turns, a very > few have more). I don't remember if we looked at every > sheet or had to quit before that; I believe we looked at > more than half. IIRC we found several dozen -- that was 17 > years ago and my memory is hazy. I have copies somewhere, > but have no interest in typing them in. Will you share all your scans so that I can type them in? > Interestingly, Dayton Miller was the first head of the > physics department at what is now CWRU. He designed their > physics building, and it has several pillars intended to > support interferometers, Why more than one? One might think he was erecting a temple ofthe Ether Wind :-) > which go all the way down to bedrock and never connect to > the rest of the building (for vibration isolation). Similar to some of my loudspeaker designs isolating the driver from the front baffle of the enclosure. > Also: the archives' exhibit on Dayton Miller was primarily > about his extensive collection of flutes; Does it include his works on acoustics, and particularly the Phonodeik? > they had to dig to find his interferometer data sheets, > which were kept loose in about a dozen folders. I believe they came from Shankland, after a long period of being unaccounted and missing. Have you an idea how I can order scans of those sheets wihtout actually coming to CWRU? If all hope for a positive result is to be abanandoned, I will do it after a diligent study of the available data, meaning some 300 or so sheets from the Mt. Wilson experiments. -- () ascii ribbon campaign -- against html e-mail /\ www.asciiribbon.org -- against proprietary attachments