fdist-methods {msbase}R Documentation

Euclidean or Manhattan distance for peak-list comparison

Description

Computes Euclidean or Manhattan distance for two aligned vectors.

Arguments

obx see above
oby see above
error measurement error
ppm if TRUE then error in parts per million, in arbitrary units otherwise
full if TRUE then compute for matching and not matching peaks
weight if TRUE weight differences between matching peaks.
uniq if TRUE compute non-crossing matching
method type of dissimilarity:
  • euclidean – euclidean distance (dist)
  • manhattan – manhattan distance. (dist)
  • canberra – canberra distance. (dist)
  • simindex – similarity index.
norm how to scale the itensities:
  • vnorm – l=2 norm, vector length normalization.
  • tic – l=1 norm, total ion count normalization.
  • student – root mean square normalization (see. scale)
  • zscore – z-score normalization. (see. scale)
theta "numeric" how to weight the non-matching peaks.
N default 0 - total length of alinged peak-lists.
range experimental

Methods

obx = "Massvector", oby = "Massvector"
:
obx = "Massvectorlist", oby = "Massvector"
:
obx = "Massvectorlist", oby = "NULL"
:

Author(s)

Witold E. Wolski witek96@users.sourceforge.net

See Also

fcor-methods,fbinary-methods

Examples

data(pldata)
fdist(pldata[[1]],pldata[[2]],error=400,ppm=TRUE,theta=1,weight=FALSE,scale="zscore",method="euclidean")
fdist(pldata[[1]],pldata[[2]],error=400,ppm=TRUE,theta=1,weight=FALSE,scale="no",method="manhattan")
tmp <- fdist(pldata,NULL,error=400,ppm=TRUE,theta=1,weight=FALSE,scale="student",method="euclidean")
plot(hclust(tmp,method="average"))

[Package msbase version 1.5.0 Index]