[R] Strange dataframe behavior

[Charles C. Berry]

On Tue, 23 Oct 2007, Sergey Goriatchev wrote:

> Hello,
>
> I have a question regarding the following output:
>
>> database <- read.delim(file=path.input.file, header=TRUE, dec=".", sep="\t", na.strings =  "#NV")
>> str(database)
> 'data.frame':   314 obs. of  13 variables:
> $ S       : Factor w/ 314 levels "307073","400212",..: 147 72 299 137
> 162 62 189 236 134 307 ...
> $ A   : Factor w/ 314 levels "Alfa",...: 285 258 197 3 81 162 183 272
> 73 301 ...
> $ M: Factor w/ 19 levels "@NA","A",..: 18 10 11 6 7 12 17 17 11 6 ...
> $ W       : num  0 0 0 0 0 ...
> $ T : num   0.0467  0.1095  0.0252  0.0821 -0.0275 ...
> $ C : num  0 0 0 0 0 ...
> $ MF   : num  -0.658  0.261  0.922 -1.897 -1.884 ...
> $ V    : num   0.0585 -1.0852 -0.3156 -1.0592  0.2810 ...
> $ G       : num  -0.568 -1.302  0.225 -1.473 -0.541 ...
> $ Mo     : num   0.34967  0.42807 -0.41407 -0.18216 -0.00305 ...
> $ R     : num  -0.5413 -2.0000  0.5353 -1.1437 -0.0776 ...
> $ Tr        : num  -0.12816  1.04148  0.00647 -0.02424 -1.66834 ...
> $ Su    : num  -1.611  1.160 -0.528 -0.091 -1.148 ...
>> which(is.na(database))
> [1] 675 704 774 887
>
> So, I have 314 observations, but there are unknown NA observations!
> I remove one observation (for certain reasons), and remove the
> corresponding factor level, then:
>> str(database)
> 'data.frame':   313 obs. of  13 variables:
> ....
>> which(is.na(database))
> [1] 673 702 772 885
>
> The removal of ONE observation moves NAs by two positions.
>
> Maybe someone have an idea what these NA observations mean????


If you mean why they move by two positions, I do.

But you can figure this out, if you do this:

 	 which( is.na(database) , arr.ind = TRUE )

and maybe this

 	row( database )
 	col( database )

Chuck

> Thanks in advance for your time and help!
>
> Sergey
> University of Zurich
>
> ______________________________________________
> 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.
>

Charles C. Berry                            (858) 534-2098
                                             Dept of Family/Preventive Medicine
E mailto:cberry at tajo.ucsd.edu	            UC San Diego
http://famprevmed.ucsd.edu/faculty/cberry/  La Jolla, San Diego 92093-0901