Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > comp.os.linux.advocacy > #116937
| From | Snit <usenet@gallopinginsanity.com> |
|---|---|
| Newsgroups | comp.os.linux.advocacy |
| Subject | Re: Long term desktop Linux trend is going up |
| Date | 2012-06-30 11:01 -0700 |
| Message-ID | <CC148CEF.4451%usenet@gallopinginsanity.com> (permalink) |
| References | (7 earlier) <e7229162-dc55-4e74-9b5d-3918f8a35621@googlegroups.com> <CC133598.434B%usenet@gallopinginsanity.com> <CC1337E5.4355%usenet@gallopinginsanity.com> <CC1354EB.437A%usenet@gallopinginsanity.com> <aedf2ada-c74d-400e-84db-72bb29047f02@t1g2000pbl.googlegroups.com> |
On 6/29/12 10:00 PM, in article aedf2ada-c74d-400e-84db-72bb29047f02@t1g2000pbl.googlegroups.com, "Onion Knight" <onionknightgot@gmail.com> wrote: > On Jun 29, 12:49 pm, Snit <use...@gallopinginsanity.com> wrote: >> There has been some debate over the long term trend of desktop Linux usage. >> >> At first I was asked to look at 24 months of data (claiming 12 months was >> "arbitrary"). So I did. And the trend line went up. I did not account for >> mathematical outliers, though, so this was deemed irrelevant. I complied >> with the request to account for outlier and found some using different >> methods (there is no one standard way, and for non-normal data the standard >> methods are not ideal, but I was told the data was close enough to normal >> for this to be OK... for the sake of this post I will accept that, though I >> really would like to see evidence). But I digress, a couple of outlier >> tests: >> >> <http://tmp.gallopinginsanity.com/Grubbs-Quartile_Nov2007-May2012.pdf> >> >> This shows the Grubbs method and the Quartile method. Neither matches the >> claims I was told had been found (including with the use of Grubbs as one of >> the methods... but no work other than mine was ever shown). >> >> I also used the median and array method: >> >> <http://tmp.gallopinginsanity.com/median_and_array.pdf> >> >> At the standard Z-Score Tolerance of 3.5 it finds no outliers, but when you >> take it to an extreme and use a Tolerance of 1 it finds 4 outliers... and >> while they do not match what I was told was to be expected, they are as >> close as any of the methods.... and it finds the most. So I went with >> that... >> >> <http://tmp.gallopinginsanity.com/long-term-trend.pdf> >> >> Page 1: Showing those four being removed >> Page 2: Showing the trend line with them removed. *Clearly* upward. >> Page 3: Repeating the same method - finding one more outlier. Even >> more importantly, though, I removed the data that has been >> shown to be largely localized to CA and this has been deemed >> anomalous *even though the mathematical models do not show >> it as such* (this is very contrary to the claims I had been >> told - no wonder he never showed his work!) >> Page 4: Even with all of *that* data removed, the trend line is >> *still* clearly moving upward, though the data is also >> not linear (if it was the R^2 value would be better and the >> fit would be better). >> >> Also consider that while the affect is likely small, the data does not >> account for the fact the overall part of the "pie" that desktop systems >> (including laptops) in general have been holding has reduced over time >> (given the growth of mobile devices). This would make the up slope >> *slightly* higher, but I suspect it is not much of a statistical difference. >> >> But there we have it: with mathematical outliers removed with a rather >> extreme process and also data known to be anomalous even outside of the >> mathematical calculations being removed, the trend for desktop Linux has >> been, overall, upward. Granted, this upward trend is slight... but there >> *is* an upward trend. >> >> I think this is excellent to see. I think as the open source community >> matures and is able to produce better systems (largely systems which act >> more as consistent systems *and* have a better selection of software) we >> will see this trend continue to go up. >> >> -- >> Summary of cc's statistical BS: <http://tinyurl.com/7rwazxw> >> Details on cc's "outliers" BS: <http://tinyurl.com/84r3ypq> >> More on cc's ignorance about outliers: <http://tinyurl.com/7vyhttc> >> On cc's claims about Grubbs and Quartiles: <http://tinyurl.com/75yvo8z> >> Details on cc's sigma and R^2 BS: <http://tinyurl.com/7vambev> > > Thank you. This is clear and focuses on the question of Linux growth > and not just on the stupid fighting. Thank you. With that said, I still look forward to cc's reply. I wish he would show how he came up with his claimed outliers (and at one time Carroll said he was able to replicate cc's work, even though it has now been shown to be absurd). Neither will ever admit they made up their claims about outliers. I just wonder what out they will try to use. :) -- Summary of cc's statistical BS: <http://tinyurl.com/7rwazxw> Details on cc's "outliers" BS: <http://tinyurl.com/84r3ypq> More on cc's ignorance about outliers: <http://tinyurl.com/7vyhttc> On cc's claims about Grubbs and Quartiles: <http://tinyurl.com/75yvo8z> Details on cc's sigma and R^2 BS: <http://tinyurl.com/7vambev>
Back to comp.os.linux.advocacy | Previous | Next — Previous in thread | Next in thread | Find similar
Re: That "Anomaly" in Linux Statistics Snit <usenet@gallopinginsanity.com> - 2012-06-28 09:59 -0700
Re: That "Anomaly" in Linux Statistics cc <scatnubbs@hotmail.com> - 2012-06-28 11:35 -0700
Re: That "Anomaly" in Linux Statistics Snit <usenet@gallopinginsanity.com> - 2012-06-28 13:53 -0700
Re: That "Anomaly" in Linux Statistics cc <scatnubbs@hotmail.com> - 2012-06-28 14:21 -0700
Re: That "Anomaly" in Linux Statistics Snit <usenet@gallopinginsanity.com> - 2012-06-28 14:38 -0700
Re: That "Anomaly" in Linux Statistics cc <scatnubbs@hotmail.com> - 2012-06-28 14:45 -0700
Re: That "Anomaly" in Linux Statistics Snit <usenet@gallopinginsanity.com> - 2012-06-28 15:16 -0700
Re: That "Anomaly" in Linux Statistics Onion Knight <onionknightgot@hotmail.com> - 2012-06-29 05:10 -0700
Re: That "Anomaly" in Linux Statistics Snit <usenet@gallopinginsanity.com> - 2012-06-29 10:36 -0700
Re: That "Anomaly" in Linux Statistics Snit <usenet@gallopinginsanity.com> - 2012-06-29 10:45 -0700
Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-29 12:49 -0700
Re: Long term desktop Linux trend is going up Big Steel <YouCanRun111277@ButYouCannotHide111277.com> - 2012-06-29 15:59 -0400
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-29 13:11 -0700
Re: Long term desktop Linux trend is going up Big Steel <YouCanRun111277@ButYouCannotHide111277.com> - 2012-06-29 16:18 -0400
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-29 13:28 -0700
Re: Long term desktop Linux trend is going up Foster <frankfoster50@yahoo.com> - 2012-06-29 16:40 -0400
Re: Long term desktop Linux trend is going up Nobody <invalid@invalid.com> - 2012-06-29 17:48 -0500
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-29 17:46 -0700
Re: Long term desktop Linux trend is going up Onion Knight <onionknightgot@gmail.com> - 2012-06-29 22:22 -0700
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 11:06 -0700
Re: Long term desktop Linux trend is going up Foster <frankfoster50@yahoo.com> - 2012-06-30 08:36 -0400
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 10:12 -0700
Re: Long term desktop Linux trend is going up Tattoo Vampire <fake@fakeemail.com> - 2012-06-30 09:55 -0400
Re: Long term desktop Linux trend is going up Nobody <invalid@invalid.com> - 2012-06-30 10:51 -0500
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 10:33 -0700
Re: Long term desktop Linux trend is going up Nobody <invalid@invalid.com> - 2012-06-30 13:04 -0500
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 11:38 -0700
Re: Long term desktop Linux trend is going up Nobody <invalid@invalid.com> - 2012-06-30 13:54 -0500
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 12:56 -0700
Re: Long term desktop Linux trend is going up Nobody <invalid@invalid.com> - 2012-06-30 15:34 -0500
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 13:56 -0700
Re: Long term desktop Linux trend is going up Nobody <invalid@invalid.com> - 2012-06-30 16:34 -0500
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 14:39 -0700
Re: Long term desktop Linux trend is going up Nobody <invalid@invalid.com> - 2012-06-30 17:05 -0500
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 18:49 -0700
Re: Long term desktop Linux trend is going up GreyCloud <mist@cumulus.com> - 2012-06-30 16:15 -0600
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 18:40 -0700
Re: Long term desktop Linux trend is going up GreyCloud <mist@cumulus.com> - 2012-06-30 22:19 -0600
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 23:16 -0700
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-29 17:36 -0700
Re: Long term desktop Linux trend is going up Foster <frankfoster50@yahoo.com> - 2012-06-30 08:26 -0400
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 10:14 -0700
Re: Long term desktop Linux trend is going up Nobody <invalid@invalid.com> - 2012-06-30 10:55 -0500
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 10:46 -0700
Re: Long term desktop Linux trend is going up Nobody <invalid@invalid.com> - 2012-06-30 13:44 -0500
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 11:58 -0700
Re: Long term desktop Linux trend is going up Nobody <invalid@invalid.com> - 2012-06-30 14:12 -0500
Re: Long term desktop Linux trend is going up Onion Knight <onionknightgot@gmail.com> - 2012-06-29 22:00 -0700
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 11:01 -0700
Re: Long term desktop Linux trend is going up Onion Knight <onionknightgot@gmail.com> - 2012-06-30 13:51 -0700
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 14:48 -0700
Re: Long term desktop Linux trend is going up Onion Knight <onionknightgot@gmail.com> - 2012-06-30 21:11 -0700
Re: Long term desktop Linux trend is going up Snit <usenet@gallopinginsanity.com> - 2012-06-30 23:09 -0700
Re: That "Anomaly" in Linux Statistics Onion Knight <onionknightgot@gmail.com> - 2012-06-29 21:59 -0700
Re: That "Anomaly" in Linux Statistics Snit <usenet@gallopinginsanity.com> - 2012-06-30 10:59 -0700
Re: That "Anomaly" in Linux Statistics Onion Knight <onionknightgot@gmail.com> - 2012-06-29 21:57 -0700
Re: That "Anomaly" in Linux Statistics High Plains Thumper <hpt@invalid.invalid> - 2012-06-29 23:02 -0600
Re: That "Anomaly" in Linux Statistics Foster <frankfoster50@yahoo.com> - 2012-06-30 08:49 -0400
Re: That "Anomaly" in Linux Statistics Snit <usenet@gallopinginsanity.com> - 2012-06-30 10:07 -0700
Re: That "Anomaly" in Linux Statistics Snit <usenet@gallopinginsanity.com> - 2012-06-30 10:55 -0700
Re: That "Anomaly" in Linux Statistics Onion Knight <onionknightgot@gmail.com> - 2012-06-29 21:41 -0700
Re: That "Anomaly" in Linux Statistics High Plains Thumper <hpt@invalid.invalid> - 2012-06-29 22:59 -0600
Re: That "Anomaly" in Linux Statistics cc <scatnubbs@hotmail.com> - 2012-07-02 04:31 -0700
Re: That "Anomaly" in Linux Statistics Snit <usenet@gallopinginsanity.com> - 2012-07-02 08:20 -0700
csiph-web