s-news
[Top] [All Lists]

Re: Probability of t

To: Trev.Parker@csiro.au
Subject: Re: Probability of t
From: Sundar Dorai-Raj <sundar.dorai-raj@pdf.com>
Date: Fri, 23 May 2003 07:06:19 -0500
Cc: s-news@lists.biostat.wustl.edu
Organization: PDF Solutions, Inc.
References: <D135ADA93CA1494295B90D3136FD93C22BC558@wallace-ah.tfrc.csiro.au>
User-agent: Mozilla/5.0 (Windows; U; Windows NT 5.0; en-US; rv:1.0.2) Gecko/20021120 Netscape/7.01


Trev.Parker@csiro.au wrote:
Hello S-plus  Fans,

I am trying to extract the test of significants i.e probability of t,  from
a GLM.  When I use the summary function eg summary(glm(formula = y ~ b3,
family = quasi(link = identity, variance = constant), data =
resample.dist)), I only get a t value for the intercept and b3, see below:

Coefficients:
Value Std. Error t value (Intercept) 0.01296489 0.0007813802 16.59229
      b3 0.46420517 0.0104191808 44.55294

How can I:
1. get the Pr(t) for the intercept and b3
2. Extract the probability as a single number into a data frame

Your help is appreciated. Thanks


Try this:

> fit <- glm(ozone^(1/3) ~ bs(radiation, 5) +
+            poly(wind, temperature, degree = 2), data = air)
> t.table <- summary(fit)$coef
> pr.t <- 2*pt(-abs(t.table[,3]),fit$df)
> t.table <- cbind(t.table, "Pr(>|t|)" = pr.t)
> data.class(t.table)
[1] "matrix"


Regards,
Sundar


<Prev in Thread] Current Thread [Next in Thread>