Skip to content

Help with loess

3 messages · eric, Pascal Oettli, Don McKenzie

#
Not sure what I'm doing wrong. Can't seem to get loess values. It looks like
loess is returning the same values as the input.

j <-loess(x1$total~as.numeric(index(x1)

plot(x1$total,type='l', ylab='M coms/y global',xlab='')
lines(loess(total~as.numeric(index(x1)),x1))

The plot statement works fine
No errors with the "lines" statement

But I don't get a smoothed value on the graph. From what I can tell, the
reason is that loess smoothed values are the same as x1$total. In other
words, x1$total -j$y = 0

What am I missing ?
AP    EU   LA    NA total
Dec 2007 3.98 14.12 6.42 26.33 50.88
Jan 2008 3.98 13.74 6.41 27.68 51.85
Feb 2008 3.90 13.80 6.39 26.07 50.18
Mar 2008 3.91 13.94 6.51 25.52 49.89
Apr 2008 3.91 14.15 6.48 25.79 50.35
May 2008 3.99 13.99 6.58 23.91 48.49
[1] "numeric"
[1] "zoo"
?zoo? series from Dec 2007 to Nov 2012
  Data: num [1:60, 1:5] 3.98 3.98 3.9 3.91 3.91 3.99 3.88 3.9 3.82 3.81 ...
 - attr(*, "dimnames")=List of 2
  ..$ : NULL
  ..$ : chr [1:5] "AP" "EU" "LA" "NA" ...
  Index: Class 'yearmon'  num [1:60] 2008 2008 2008 2008 2008 ...



--
View this message in context: http://r.789695.n4.nabble.com/Help-with-loess-tp4650132.html
Sent from the R help mailing list archive at Nabble.com.
#
Hello,

Your problem is not reproducible.

Regards,
Pascal


Le 20/11/2012 12:20, eric a ?crit :
#
In your case j$y are the original Y values (response), not the fitted  
values.   Try

plot(j$x,fitted(j))
On 19-Nov-12, at 7:20 PM, eric wrote:

            
Don McKenzie, Research Ecologist
Pacific Wildland Fire Sciences Lab
US Forest Service
phone: 206-732-7824

Affiliate Professor
School of Environmental and Forest Sciences
University of Washington