I am using R version of 2.15.3 in R studio environment. I used a function "skew" but got an error message that the name function cannot be found.
You know how skew and kurtosis are calculated (or look it up, e.g. in Wikipedia). So you can write your functions.
Note that different formulas for computing skewness and kurtosis can be used. See, for example, the options "type" of the function kurtosis() shown by using ?kurtosis() after loading the package e1071 by using library(e1071).
You can use package 'moments'. Check following link to get pdf of this package
http://cran.r-project.org/web/packages/moments/moments.pdf
Good luck!
You can also use the library "propagate". Install it in R, load it and use the functions kurtosis() and skewness() as follows:
> library (propagate)
> X skewness(X)
> kurtosis(X)
The psych package has a function describe(), which gives quite a few descriptives including skewness and kurtosis.
I believe it uses the methods similar to e1071, but you might want to confirm that.
This link can help you
https://alstatr.blogspot.com/2013/06/measures-of-skewness-and-kurtosis.html
I see that Jully Ouma has already mentioned the moments package. This is nice because it outputs a single figure. For a range of diagnostic and summary statistics, you could use describe() from the psych package, or if you want more detail you could try stat.desc() from the pastecs package. This is the most comprehensive of the options I think.
Using the R's fbasic Package: {skewness- returns value of skewness;
kurtosis- returns value of kurtosis}
USAGE:
## simple method:
skewness (x, ...)
## And the Default S3 method:
skewness(x, na.rm = FALSE, method = c("moment", "fisher"), ...)
## S3 method for class 'data.frame':
skewness(x, ...)
## S3 method for class 'POSIXct':
skewness(x, ...)
## S3 method for class 'POSIXlt':
skewness(x, ...)
KURTOSIS{USAGE}
##simple method:
kurtosis(x, ...)
## Default S3 method:
kurtosis (x, na.rm = FALSE, method = c("excess", "moment", "fisher"), ...)
## S3 method for class 'data.frame':
kurtosis(x, ...)
## S3 method for class 'POSIXct':
kurtosis(x, ...)
## S3 method for class 'POSIXlt':
kurtosis(x, ...).
NOTE_PLS:- the syntax "Method" is a character string which specifies the method of computation. These are either "moment" or "fisher". {however, kurtosis allows in addition for "excess". i.e., If "excess" is selected, then the value of the kurtosis is computed by the "moment" method and a value of 3 will be subtracted.}
a) The "moment" method is based on the definitions of skewness and kurtosis for distributions; these forms should be used when resampling (bootstrap or jackknife).
b) The "fisher" method correspond to the usual "unbiased" definition of sample variance, although in the case of skewness and kurtosis exact unbiasedness is not possible.
**Aliter: other packages exist, of which one are "moments", "propagate" and "psych".
EXample:
*From the psych package.
>library(psych)
>skew(file$factorA)
>kurtosi(filename$factorA)
:)
Goodluck.
One possible procedure is describe as follows:
> install.packages("moments")
##after installation select moments from packages or simply use
> library(moments)
> x skewness(x)
[1] 0.1181213
- Badges
- Science topic