[R] read floats from file into array

[Duncan Murdoch]

On 21/07/2009 6:09 AM, leo mueller wrote:
> hi all,
> 
> i have a simple question. instead of defining my measurements in a
> static way like ...
> 
> x <- c(-0.475, -1.553, -0.434, -1.019, 0.395)
> 
> ... i'd like them to be read from a file ...
> 
> x <- read.table("07a673ac0cb1f7f8fa293860566f633c/1/raw0.txt", header=FALSE)
> d1 <- density(x, kernel = "gaussian")
> 
> with a formatting that looks like:
> 
> 4.2840000000e-01
> 6.7583333333e-01
> 8.2920000000e-01
> 7.8566666667e-01
> 6.6336666667e-01
> 5.4080000000e-01
> 4.7283333333e-01
> 4.3770000000e-01
> 4.3743333333e-01
> 4.1026666667e-01
> 3.6283333333e-01
> 3.2770000000e-01
> 4.9096666667e-01
> [...]
> 
> R quits and says:
> 
>> d1 <- density(x, kernel = "gaussian")
> Error in density.default(x, kernel = "gaussian") :
>   argument 'x' must be numeric
> Calls: density -> density.default
> Execution halted
> 
> is there any possibility to convert this / make this work?

read.table returns a dataframe, i.e. a list of vectors.  density wants a 
vector.  So you will probably get what you want using

d1 <- density(x[[1]], kernel="gaussian")

You can use names(x) to find the name of the 1st column for a nicer 
syntax; it is probably V1 (for "variable 1"), so you could do

y <- x$V1
density(y, ...)

Duncan Murdoch