R help - Aug 2011 - Wilcox test

Hello,

I have used Wilcox test to find the p-value for hgua95 spiken data. It did
not give the required result, and instead gave me warnings like the values
are duplicated. Kindly suggest me how to overcome this problem as I am new
to R.


Thank you

Deepthi BM

<bdeepthi <at> ibab.ac.in> writes:
> 
> Hello,
> 
> I have used Wilcox test to find the p-value for hgua95 spiken data. It did
> not give the required result, and instead gave me warnings like the values
> are duplicated. Kindly suggest me how to overcome this problem as I am new
> to R.
  Please read the posting guide (and <http://tinyurl.com/3jgynql>)
and give us a reproducible example.
We don't know what the "hgua95 spiken (?) data" is; we don't
know what
the "required result" is; presumably if you got warnings about
duplicated
values, then you had duplicated values in your data, either because
they were really there or because the data got imported wrong or because
it got mangled after you imported it ...

Hi,

Two queries, if I may ask:

A. I wish to retrieve the p-values from wilcox test for EACH gene  --- such
that Samples 1-3 is considered as one group while the rest remains in the
other group.

There is 1013 genes in the dataframe as attempted to exemplify in the table
below.


                    |Sample 1 | Sample2 | sample3 | SDample4 | sample 5 |
sample 6
_________________________________________________________________________
Gene 1        | 2.4          | 3.5          | 4.2          | 6.2
| 2.4          | 6.3
Gene 2        | 6.4          | 6.5          | 2.2          | 8.2
| 4.4          | 2.3
----
-----
Gene 1013  | 1.4          | 2.5          | 3.2          | 4.2            |
5.4           | 7.3


Accordingly, the following function is made.

funn <- function (x,y)
{

for (i in 1:1013)
  {

 wil <- wilcox.test(as.numeric(x), as.numeric (y), paired=TRUE);
 print (wil);
  }

return (wil);

}


However, I think it is somewhat incorrect as commands like the following
generates several warnings (and it gives more than just the p-value).


funn (c(file [,1]:file [,3]), c(file[,4]:file [,6])) # "file" is the
dataframe.


B. Any good books/weblink you could advise which focus more on the
programming side of r, such as function-building ? The few books that I have
gone thorugh have very less on such aspect - they focus more on statistical
aspects and use of built-in functions.

Thank you.

Cheers,
Chintanu

	[[alternative HTML version deleted]]