R help - Jun 2013 - Unexpected behavior from hist()

Hi... I'm still a beginner in R. While doing some curve-fitting with a raw
data set of length 22,000, here is what I had:


> hist(y,col="red")
gives me the frequency histogram, 13 total rectangles, highest is near 5000.



Now
> hist(y,prob=TRUE,col="red",ylim=c(0,1.5))


gives me the density (probability?) histogram, same number f rectangles, but the
highest rectangle is obviously higher than 1, how can this be?!!!



And it gets worse, if I use the 'breaks' option, when I add
'breaks=1000', many of the rectangles were higher than 2.

What am I missing here?



Thanks.


P.S. I had to post this thread via email as it got rejected as I posted it from
Nabble, reason was "Message rejected by filter rule match"


	[[alternative HTML version deleted]]

Hi,

On Thu, Jun 13, 2013 at 11:13 AM, Mohamed Badawy <mbadawy at pm-engr.com>
wrote:
> Hi... I'm still a beginner in R. While doing some curve-fitting with a
raw data set of length 22,000, here is what I had:
>
>
>
>> hist(y,col="red")
>
> gives me the frequency histogram, 13 total rectangles, highest is near
5000.
>
You don't provide a reproducible example, so here's some fake data:

somedata <- runif(1000)

> Now
>
>> hist(y,prob=TRUE,col="red",ylim=c(0,1.5))
>
> gives me the density (probability?) histogram, same number f rectangles,
but the highest rectangle is obviously higher than 1, how can this be?!!!
Because you misread the help. using freq=FALSE (equivalent to
prob=TRUE, which is a legacy option), you are getting:

freq: logical; if ?TRUE?, the histogram graphic is a representation
          of frequencies, the ?counts? component of the result; if
          ?FALSE?, probability densities, component ?density?, are
          plotted (so that the histogram has a total area of one).
          Defaults to ?TRUE? _if and only if_ ?breaks? are equidistant
          (and ?probability? is not specified).


It sounds like what you actually want is:

somehist <- hist(somedata, plot=FALSE)
somehist$counts <- somehist$counts/sum(somehist$counts)
plot(somehist)
> P.S. I had to post this thread via email as it got rejected as I posted it
from Nabble, reason was "Message rejected by filter rule match"
Nabble is not the R-help mailing list. Posting via email is the
correct thing to do.

Sarah

-- 
Sarah Goslee
http://www.functionaldiversity.org