Ok,many thanks for your detailed answer.
At 2013-09-18 19:20:26,"Duncan Murdoch" <murdoch.dun...@gmail.com> wrote: >On 13-09-18 1:38 AM, meng wrote: >> Oh,yes, I found out this according to your reply.Thanks. >> >> As to time series analysis, in order to show the effect of smoothing or >> filtering,the common command is: >> plot(ts0); >> lines(fitted(...)) >> But not "lines(fitted(...) ~ time(ts) )" >> >> How to understand this then? > >lines() and plot() are "generic functions". What it does depends on the >class of the first argument. To see what happens, you need to know the >class of ts0, or fitted(...), or fitted(...) ~ time(ts). I'd guess ts0 >has some time series class, fitted(...) probably has class "numeric" >(though this would depend on the dots, since it is also generic), and >the formula has class "formula". "numeric" generally gets the default >method (plot.default, lines.default); "formula" usually has its own >methods (plot.formula, lines.formula), etc. Read up on this in An >Introduction to R for more details (sections 3.4 and 10.9). > >Duncan Murdoch > >> >> Many thanks. >> >> Best. >> >> >> >> >> At 2013-09-18 08:49:51,"Duncan Murdoch" <murdoch.dun...@gmail.com> wrote: >>>On 13-09-17 6:36 PM, meng wrote: >>>> Thanks for your reply. >>>> >>>> Is "fitted(lm(...))" the same as "values" of lines(values)? >>>> >>>> If yes,then why the range of lines(values) is different from >>>> range(fitted(lm(...)))? >>> >>>You are plotting against the wrong x axis, and you don't see all the values. >>> >>>Duncan Murdoch >>> >>>> If no, what "values" refers to? >>>> >>>> >>>> >>>> At 2013-09-17 20:56:04,"Duncan Murdoch" <murdoch.dun...@gmail.com> wrote: >>>>>On 13-09-17 8:06 AM, meng wrote: >>>>>> Hi all: >>>>>> I met a question about "lines". >>>>>> >>>>>> >>>>>> attach(cars) >>>>>> >>>>>> >>>>>> plot(dist ~ speed) >>>>>> #add the regression line to the plot >>>>>> lines(fitted(lm(dist~speed)) ~ speed) >>>>>> >>>>>> >>>>>> plot(dist ~ speed) >>>>>> #what kind of curve does the following command add to the plot? >>>>>> lines(fitted(lm(dist~speed))) >>>>>> >>>>>> >>>>>> My question is : >>>>>> what kind of curve does the last command add to the plot? >>>>> >>>>>Look at the class of fitted(lm(...)). It is "numeric". So what you're >>>>>seeing is the same as if you computed the fitted values, and then did >>>>> >>>>>lines(values) >>>>> >>>>>Since values is just a vector of numbers, that will plot them as y >>>>>values against x values 1:length(values). That's unlikely to be a >>>>>useful thing to do. >>>>> >>>>>Duncan Murdoch >>>>> >>>>>> >>>>>> >>>>>> My guess:maybe the level of fitted values? >>>>>>> range(fitted(lm(dist~speed))) >>>>>> [1] -1.84946 80.73112 >>>>>> >>>>>> >>>>>> But from the plot,I can see the range of the curve is about 10 to 40 >>>>>> more or less,which is different from(-1.84946, 80.73112).So the curve >>>>>> must not be the fitted values.What kind of curve does the last command >>>>>> add to the plot then? >>>>>> >>>>>> >>>>>> >>>>>> >>>>>> Many thanks for your help >>>>>> >>>>>> >>>>>> My best >>>>>> [[alternative HTML version deleted]] >>>>>> >>>>>> ______________________________________________ >>>>>> R-help@r-project.org mailing list >>>>>> https://stat.ethz.ch/mailman/listinfo/r-help >>>>>> PLEASE do read the posting guide >>>>>> http://www.R-project.org/posting-guide.html >>>>>> and provide commented, minimal, self-contained, reproducible code. >>>>>> >>>>> >>>> >>>> >>>> >>> >> >> >> > [[alternative HTML version deleted]] ______________________________________________ R-help@r-project.org mailing list https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide http://www.R-project.org/posting-guide.html and provide commented, minimal, self-contained, reproducible code.
