I belive that the problem is not with the table, but with your predictions
which are not 0s and 1s.
Ales Ziberna
- Original Message -
From: "Stephen Choularton" <[EMAIL PROTECTED]>
To: "'R Help'"
Sent: Wednesday, April 27, 2005 4:19 AM
Subject: [R]
I am trying to do some verification across a large dataset, cuData, that
has 23 columns.
Column 23 (similarity) is the outcome 0 or 1 and the other columns are
the features.
I do this:
verificationglm.model <- glm(formula = similarity ~ ., family=binomial,
data=cuData[1:1000,])
and produce