R help - Jul 2003 - Probability plotting with R

Hello,

Our professor asked us to do probability plotting using weibull paper,
exponential paper, normal, log-normal paper, etc. I know I can create Q-Q
plot for normal dist. and see if all te points are on one line. How do I
go about other distributions?

I tried generating different samples and use the general qq function.
However, I could not do it since I don't know the population parameters
(In normal QQ, we assume that the sample mean and variance were population
parameters too)

I was wondering if there is any way of probability plotting with R instead
of using the papers. I appreciate your helps in advance.

Regards

A. Gunes Koru

On Fri, 18 Jul 2003, A. Gunes Koru wrote:
> Hello,
>
> Our professor asked us to do probability plotting using weibull paper,
> exponential paper, normal, log-normal paper, etc. I know I can create Q-Q
> plot for normal dist. and see if all te points are on one line. How do I
> go about other distributions?
>
The qqmath() function in the lattice package can do plots like those from
qqnorm() for other distributions.

	-thomas

On Fri, 18 Jul 2003 08:50:01 -0700 (PDT), you wrote:
>On Fri, 18 Jul 2003, A. Gunes Koru wrote:
>
>> Hello,
>>
>> Our professor asked us to do probability plotting using weibull paper,
>> exponential paper, normal, log-normal paper, etc. I know I can create
Q-Q
>> plot for normal dist. and see if all te points are on one line. How do
I
>> go about other distributions?
>>
>
>The qqmath() function in the lattice package can do plots like those from
>qqnorm() for other distributions.
Also qqplot used with ppoints, e.g.

# make up some fake data
x <- rexp(100,10)

# calculate standard quantiles to plot against
stdexponential <- qexp(ppoints(100))

# do the plot
qqplot(stdexponential,x)