[R] Fitting a curve to data

Carl Witthoft carl at witthoft.com
Wed Jul 2 01:07:35 CEST 2008

And, oddly enough :-),  integrate.xy does pretty much exactly what I 
suggested.  Thanks for providing that reference

I would  be interested in seeing how the original poster's data works 
out using integrate.xy as opposed to simply calculating x*y

By the way, since the original data were 'percent of total volume' over 
each hour of the day,  how could the integral NOT be 100%?


Gabor Grothendieck wrote:
> There is an integrate.xy in sfsmic.  Limitations discussed there.
> On Tue, Jul 1, 2008 at 6:27 PM, stephen sefick <ssefick at gmail.com> wrote:
>> I would like to know the answer to this question now that I know what we are
>> getting at.  integrate() looks like it is the right thing, but it has to use
>> a function-  I would like to know how to just integrate the area under a
>> curve with just an input of x and y coordinates.
>> Stephen
>> On Tue, Jul 1, 2008 at 5:22 PM, Carl Witthoft <carl at witthoft.com> wrote:
>>> I think the previous answer (to use lm() ) is not necessarily the best
>>> option.
>>> Since what you want is the definite integral (area under the curve), you
>>> can just use one of the existing definite integration tools (sorry, I don't
>>> recall the names because I don't use them).
>>> If you want to get a "smoothed" curve to remove errors in your data points,
>>>  I'd recommend fitting a spline and integrating under that.

More information about the R-help mailing list