Rank ordering xts objects representing some property of equities using R

I am trying to rank order equities (by return for example). As a result I would like to receive a table containing names of stocks in ascending/descending order (parameter to this rank order function) with proper handling of NAs (moved at the end of each row). I really can't figure out an elegant way to do this.

Below is an example of what I want:

This is coredata of xts object representing some property at different times:

           john    joe     tina    jack    suzie   sasha   sven    luca   
2003-05-29 1 开发者_Python百科      2       3       4       5       6       7       8  
2003-06-27 2       3       4       5       6       7       8       1  
2003-07-30 3       4       5       6       7       8       1       2
2003-07-31 NA      2       3       4       5       6       1       NA

I need a dataframe which in each row (for each date) displays the column name from previous dataframe of best ranked (based on a property) in column 1, second best in column 2,... Please note that for the last row I need cases with NAs moved to the end (last two columns) or skipped...

           [,1]    [,2]    [,3]    [,4]    [,5]    [,6]    [,7]    [,8]   
2003-05-29 "john"  "joe"   "tina"  "jack"  "suzie" "sasha" "sven"  "luca"  
2003-06-27 "luca"  "john"  "joe"   "tina"  "jack"  "suzie" "sasha" "sven"
2003-07-30 "sven"  "luca"  "john"  "joe"   "tina"  "jack"  "suzie" "sasha"
2003-07-31 "sven"  "joe"   "tina"  "jack"  "suzie" "sasha" "john"  "luca" 

Thanks in advance for your help. Being a beginner in R this poses to hard of a problem for me...

Kind regards,

Samo.

Suppose m is your original matrix (i.e. the coredata of the xts). You can then get what you want with:

> nams <- colnames(m)
> t( apply(m, 1, function(r) nams[ order(r) ] ) )
           [,1]   [,2]   [,3]   [,4]   [,5]    [,6]    [,7]    [,8]   
2003-05-29 "john" "joe"  "tina" "jack" "suzie" "sasha" "sven"  "luca" 
2003-06-27 "luca" "john" "joe"  "tina" "jack"  "suzie" "sasha" "sven" 
2003-07-30 "sven" "luca" "john" "joe"  "tina"  "jack"  "suzie" "sasha"
2003-07-31 "sven" "joe"  "tina" "jack" "suzie" "sasha" "john"  "luca" 

Note that order(r) converts a vector r to a vector p such that the sequence r[p[1]], r[ p[2] ], r[ p[3] ], ... is in non-decreasing order, and the default is to push NAs to the end.