0
$\begingroup$

I am taking random numbers from an uniform distribution. After I have taken some number of samples, I plot the probability density function. Now, I'm wondering how should I call the plot. Is it probability density function of uniform sampling distribution? Due to small number of samples it does not look like exactly uniform distribution, so I don't know what would be the correct term. Near-uniform?

Edit:

x = rand(1000, 1);
[f, t] = ksdensity(x);
figure;
plot(t, f);
$\endgroup$
5
  • $\begingroup$ Exactly how are you are producing a plot? This affects what terminology is advisable. $\endgroup$
    – Nick Cox
    Aug 23, 2013 at 13:20
  • $\begingroup$ Uhm... a histogram? $\endgroup$
    – StasK
    Aug 23, 2013 at 13:22
  • $\begingroup$ If you are taking samples from a uniform distribution and plotting the PDF then you can call it the PDF from a uniform sample. Providing the $\textit{n}$ may be helpful. $\endgroup$ Aug 23, 2013 at 13:23
  • $\begingroup$ x = rand(1000, 1); [f, t] = ksdensity(x); figure; plot(t, f); $\endgroup$
    – asda
    Aug 23, 2013 at 13:24
  • 3
    $\begingroup$ Looks like MATLAB to me. "Say what programming language you're using." meta.stats.stackexchange.com/questions/1479/… $\endgroup$
    – Nick Cox
    Aug 23, 2013 at 13:58

1 Answer 1

3
$\begingroup$

On the information to date, I suggest some wording like

Kernel estimate of probability density function of uniform distribution based on a sample of 1000 from rand(1000,1) in MATLAB.

Note that default procedures don't usually do a good job at the edges of a bounded probability distribution.

Another terminology problem: I think most statistical people would say that you have one sample of size 1000. "number of samples" suggests something quite different in statistical science. This contrasts with usages in several scientific and practical disciplines in which each measurement is regarded as one sample. (This has been discussed elsewhere on CV: I will try to find the thread.)

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge that you have read and understand our privacy policy and code of conduct.

Not the answer you're looking for? Browse other questions tagged or ask your own question.