[R] Probit Analysis

[apjaworski@mmm.com]

Jacquelin,

Here is an answer to your question wrapped in much longer question of my
own.

Cheers,

Andy

-----------------------------------------------------------------------------------------

Extracting components from complex objects is, in my opinion, not easy to
do unless I am missing something very obvious.  Take the glm.dose object as
an example.  Following the example on the dose.p help page I ran

      zz <- dose.p(budworm.lg0, cf = c(1,3), p = 1:3/4)

Typing zz applies print method to the zz object and shows

                          Dose        SE
      p = 0.25: 2.231264 0.2498235
      p = 0.50: 3.263587 0.2297039
      p = 0.75: 4.295910 0.2746190

If I now want to extract some of the numbers from zz, I have to go through
some hoops.  What I usually do first is

      str(zz)

This gives me

      Class 'glm.dose'  atomic [1:3] 2.23 3.26 4.30
        ..- attr(*, "SE")= num [1:3, 1] 0.250 0.230 0.275
        .. ..- attr(*, "dimnames")=List of 2
       .. .. ..$ : chr [1:3] "p = 0.25:" "p = 0.50:" "p = 0.75:"
        .. .. ..$ : NULL
        ..- attr(*, "p")= num [1:3] 0.25 0.5 0.75

The format of this output is not really described on the str help page, so
I am guessing that what this means is that the "basic" object is a vector
of length 3 with a couple of attributes called SE and p, the SE attribute
having additional nested attribute called dimnames.  We can onfirm that by

      attributes(zz)

giving

      $names
      [1] "p = 0.25:" "p = 0.50:" "p = 0.75:"

      $SE

                   [,1]
       p = 0.25: 0.2498235
       p = 0.50: 0.2297039
       p = 0.75: 0.2746190

      $p
      [1] 0.25 0.50 0.75

      $class
      [1] "glm.dose"

So, If I want to extract the actual dose value from zz I just need

      zz[1]

geting

      p = 0.25:
      2.231264

but if I want to get the first standard error I have to do

      attributes(zz)$SE[1]

getting

      p = 0.25:
      0.2498235


Is this a correct general method of extracting components from complex
objects?  Is there a simpler way?

Thanks in advance and sorry for this longish post, but this is something I
have been having problems with for a long time and it seems that every time
I run into this "extraction problem" I have to reinvent the wheel.

Andy

__________________________________
Andy Jaworski
Engineering Systems Technology Center
3M Center, 518-1-01
St. Paul, MN 55144-1000
-----
E-mail: apjaworski at mmm.com
Tel:  (651) 733-6092
Fax:  (651) 736-3122


|---------+---------------------------------->
|         |           "Ngayee J Law"         |
|         |           <Ngayee.Law at celeradiagn|
|         |           ostics.com>            |
|         |                                  |
|         |           12/31/2002 17:18       |
|         |                                  |
|---------+---------------------------------->
  >-----------------------------------------------------------------------------------------------------------------------------|
  |                                                                                                                             |
  |      To:       apjaworski at mmm.com                                                                                           |
  |      cc:       r-help at stat.math.ethz.ch                                                                                     |
  |      Subject:  Re: [R] Probit Analysis                                                                                      |
  >-----------------------------------------------------------------------------------------------------------------------------|




Hi Andy,

Thanks! glm and dose.p can do exaclty what I want to do. Just a quick
question,
how can I extract the numbers in the glm.dose object created by the dose.p?
My
goal is to calculate a 95% CI for the 'LD50'.

- Jacqueline





                    apjaworski at mm

                    m.com                To:     "Ngayee J Law"
<Ngayee.Law at celeradiagnostics.com>
                                         cc:     r-help at stat.math.ethz.ch

                    12/31/2002           Subject:     Re: [R] Probit
Analysis
                    02:47 PM








Jacqueline,

This should be a simple application of GLM.  Check the help pages for glm,
family and dose.p in library MASS.  Look at the example at the bottom of
the dose.p help page - it does almost exactly what you need.

Hope this helps,

Andy

__________________________________
Andy Jaworski
Engineering Systems Technology Center
3M Center, 518-1-01
St. Paul, MN 55144-1000
-----
E-mail: apjaworski at mmm.com
Tel:  (651) 733-6092
Fax:  (651) 736-3122


|---------+---------------------------------->
|         |           "Ngayee J Law"         |
|         |           <Ngayee.Law at celeradiagn|
|         |           ostics.com>            |
|         |           Sent by:               |
|         |           r-help-admin at stat.math.|
|         |           ethz.ch                |
|         |                                  |
|         |                                  |
|         |           12/31/2002 16:17       |
|         |                                  |
|---------+---------------------------------->
  >
-----------------------------------------------------------------------------------------------------------------------------|


  |
|
  |      To:       r-help at stat.math.ethz.ch
|
  |      cc:
|
  |      Subject:  [R] Probit Analysis
|
  >
-----------------------------------------------------------------------------------------------------------------------------|





Hello all,

I have a very simple set of data and I would like to analyze them with
probit analysis.
The data are:

X    Event          Trial
100  8         8
75   8         8
50   6         8
25   4         8
10   2         8
0    0         8

I want to estimate the value of X that will give a 95% hit rate
(Event/Trial) and the corresponding
95% CI. Anyone can offer some help? Thanks!!

- Jacqueline

______________________________________________
R-help at stat.math.ethz.ch mailing list
http://www.stat.math.ethz.ch/mailman/listinfo/r-help