Odp: log2() and -min() very quick question

Hi

r-help-bounces at r-project.org napsal dne 13.06.2011 17:59:03:
Ben Ganzfried <ben.ganzfried at gmail.com> 
Odeslal: r-help-bounces at r-project.org

13.06.2011 17:59

Komu

r-help at r-project.org

Kopie

P?edm?t

[R] log2() and -min() very quick question

I'm looking over good-code a post-doc in my lab wrote and trying to 
learn
how it works.  I came across the following:
rel.abundance <- 
as.matrix(read.delim("rel.abundance.csv",row.names=1,as.is
=TRUE))
rel.abundance <- log2(rel.abundance-min(rel.abundance)+1)

I'm not sure what the second line is doing.  I ran each line in R and
couldn't see a noticeable difference in the output.  I assume log2() 
takes
the log base 2 of the values?  I'm not clear what -min(rel.abundance) is
doing either...my hunch would be that it would take the smallest value 
in
each row?
No. If rel.abundance is matrix min(rel.abundance) is overall minimum
mat<-matrix(1:12, 3,4)
min(mat)
[1] 1

so
log2(rel.abundance-min(rel.abundance)+1)

subtract minimum value from all numbers, after that it add 1 do all 
numbers, takes log base 2 from each number and returns matrix with the 
same dimensions as input matrix.
I'd really like to figure out:
1) What's actually going on?
2) Is there a good way to run a command over a large dataset in R and 
better
be able to tell what is going on?  More specifically, when I run each 
line
in R it looks something like this (w/ dif. values per row):
Archaea|Euryarchaeota|Methanobacteria|Methanobacteriales|

Methanobacteriaceae|Methanobrevibacter,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,

0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,

0,0,0,0,0,0,0,0,0,3,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,

0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,3,0,0,0,0,0,

0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,

0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,2,
0,0,0,0,0,0,0,0,0,1,0,0,0,0,0,0,0,0,0,0,0,0,0,23,0,3,0,0,0

There are a lot of cells w/ values per row, which is one reason why I 
think
it is difficult to detect a pattern....
there are some summary and structure commands

summary(data) or str(data)

which can tell you some overall information about your data.

Regards
Petr
Thanks in advance!

Ben

   [[alternative HTML version deleted]]

______________________________________________
R-help at 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.

Odp: log2() and -min() very quick question

Thread (3 messages)