An embedded and charset-unspecified text was scrubbed... Name: not available URL: <https://stat.ethz.ch/pipermail/r-help/attachments/20121220/82b4cf3b/attachment.pl>
how to make a table of summary statistics
3 messages · Francesco Sarracino, Duncan Murdoch, Jim Lemon
On 12-12-20 6:45 AM, Francesco Sarracino wrote:
Dear R-listers,
I am a newbie with R and I am struggling with something I consider very
basic. I wish to produce a table (to import in a latex file) of summary
statistics, but for as much as I've been looking around and trying various
alternatives (plyr, reporttools, pastecs and Hmisc) I haven't found what I
am looking for. Probably I am doing something wrong, but I can't figure out
what.
Let's make up three simple variables:
var1 <- runif(1000)
var2 <- runif(1000)
var3 <- factor(rep(1:2, 500), labels = c("m", "f"))
and let's create a dataset out of them:
data <- data.frame(var1, var2, var3)
what I'd like to get is a table such as the following one:
variable mean sd min max obs missing
var1
var2
var3
where for each variable, I can read in line the mean, the standard
deviation, the min and the max values, the number of observations and the
percentage of missing data.
Can you advice any way to achieve it?
Thanks a lot in advance for your kind help,
I'm not sure what you want for var3: it doesn't make sense to calculate
the mean or sd for a factor. But for the other variables, using package
tables, you do
latex( tabular( Heading("variable")*(var1 + var2) ~ (mean + sd + min +
max + (obs=length) + (missing=function(x) sum(is.na(x)))), data=data) )
You might want a breakdown of the summaries by var3; you'd get that this
way:
latex( tabular( Heading("variable")*(var1 + var2)*var3 ~ (mean + sd +
min + max + (obs=length) + (missing=function(x) sum(is.na(x)))),
data=data) )
Duncan Murdoch
On 12/20/2012 10:45 PM, Francesco Sarracino wrote:
Dear R-listers,
I am a newbie with R and I am struggling with something I consider very
basic. I wish to produce a table (to import in a latex file) of summary
statistics, but for as much as I've been looking around and trying various
alternatives (plyr, reporttools, pastecs and Hmisc) I haven't found what I
am looking for. Probably I am doing something wrong, but I can't figure out
what.
Let's make up three simple variables:
var1<- runif(1000)
var2<- runif(1000)
var3<- factor(rep(1:2, 500), labels = c("m", "f"))
and let's create a dataset out of them:
data<- data.frame(var1, var2, var3)
what I'd like to get is a table such as the following one:
variable mean sd min max obs missing
var1
var2
var3
where for each variable, I can read in line the mean, the standard
deviation, the min and the max values, the number of observations and the
percentage of missing data.
Hi Francesco, Have a look at the describe function (prettyR) that allows you to select the summary statistics you want for numeric variables and automatically produces count statistics for factor, character and logical variables. There are also other "auto-summarize" functions in packages such as psych and Hmisc. Jim