Benjamin Cheah
2010-Jan-25 01:08 UTC
[R] Functional data analysis - problem with functional linear regression
Hi all, I'm trying to perform a functional linear regression with functional response (comprising 14 coordinates) and 4 scalar covariates. When I create a list containing the scalar covariates (with 14 replications each), I keep getting the following message: [1] "Vector in XFDLIST[[ 1 ]] has wrong length." [1] "Matrix in XFDLIST[[ 1 ]] has more than one column." [1] "Vector in XFDLIST[[ 2 ]] has wrong length." [1] "Matrix in XFDLIST[[ 2 ]] has more than one column." [1] "Vector in XFDLIST[[ 3 ]] has wrong length." [1] "Matrix in XFDLIST[[ 3 ]] has more than one column." [1] "Vector in XFDLIST[[ 4 ]] has wrong length." [1] "Matrix in XFDLIST[[ 4 ]] has more than one column." Error in fRegress.fdPar(PPIfdPar, predictor, betalist) : problem with xfdlist So in contrast to what is advised in the R FDA manual, I reduce my scalar covariate matrices to column vectors with 88 rows (for 88 indepedent observations), each. This message pops up: [1] "Vector in XFDLIST[[ 1 ]] has wrong length." [1] "Vector in XFDLIST[[ 2 ]] has wrong length." [1] "Vector in XFDLIST[[ 3 ]] has wrong length." [1] "Vector in XFDLIST[[ 4 ]] has wrong length." Error in fRegress.fdPar(PPIfdPar, predictor, betalist) : problem with xfdlist Can anyone offer any solutions? Many thanks in advance! Ben Cheah __________________________________________________________________________________ See what's on at the movies in your area. Find out now: http://au.movies.yahoo.com/session-times/ [[alternative HTML version deleted]]