[R] Chisq test for truncated count data and estiamte the expectation value
pinusan
anhong at msu.edu
Thu Feb 18 02:50:58 CET 2010
Dear all,
I have some questions for count data.
First, I want to calculate the expectation value and variance from truncated
data when I assume that the truncated data follow Poisson or negative
binomial distribution.
For example,
N freq
1 30
2 35
3 26
4 8
5 0
6 2
7 0
The mean of N is 2.2. If the data follows Poisson distribution, the
estimated lambda is 1.9.
Is there any package to estimate the lambda from truncated count data?
I used “gnlm” and “vcd” packages but I did not have correct expectation
value.
Second, I want to test the goodness of fit using truncated count data(see.
example data) for the given distributions such as Poisson or negative
binomial distribution. I think Chisq test can be one solution to test the
goodness of fit. However, I need to run more than 10,000 data sets. I
think it takes long time to test the goodness of fit. Therefore, I need to
find a method which can test the goodness of fit of truncated count data for
the given distribution.
If you have any suggestions (packages, book, or paper), it will be really
helpful to my research.
Have a nice day.
--
View this message in context: http://n4.nabble.com/Chisq-test-for-truncated-count-data-and-estiamte-the-expectation-value-tp1559590p1559590.html
Sent from the R help mailing list archive at Nabble.com.
More information about the R-help
mailing list