Probability of familial clustering of disease

Usage

pfc(famdata, enum = 0)

Arguments

famdata: collective information of sib size, number of affected sibs and their frequencies.
enum: a switch taking value 1 if all possible tables are to be enumerated.

Value

The returned value is a list containing (tailp,sump,nenum are only available if enum=1:

p the probabitly of familial clustering.
stat the deviances, chi-squares based on binomial and hypergeometric distributions, the degrees of freedom should take into account the number of marginals used.
tailp the exact statistical significance.
sump sum of the probabilities used for error checking.
nenum the total number of tables enumerated.

Details

To calculate exact probability of familial clustering of disease

Note

Adapted from family.for by Dani Zelterman, 25/7/03

References

Yu C, Zelterman D (2001). “Exact inference for family disease clusters.” Communications in Statistics - Theory and Methods, 30(11), 2293-2305. doi:10.1081/STA-100107686 .

Yu C, Zelterman D (2002). “Statistical inference for familial disease clusters.” Biometrics, 58(3), 481-91. doi:10.1111/j.0006-341x.2002.00481.x .

Author

Dani Zelterman, Jing Hua Zhao

Examples

if (FALSE) { # \dontrun{
# IPF among 203 siblings of 100 COPD patients from Liang KY, SL Zeger,
# Qaquish B. Multivariate regression analyses for categorical data
# (with discussion). J Roy Stat Soc B 1992, 54:3-40

# the degrees of freedom is 15
famtest<-c(
1, 0, 36,
1, 1, 12,
2, 0, 15,
2, 1,  7,
2, 2,  1,
3, 0,  5,
3, 1,  7,
3, 2,  3,
3, 3,  2,
4, 0,  3,
4, 1,  3,
4, 2,  1,
6, 0,  1,
6, 2,  1,
6, 3,  1,
6, 4,  1,
6, 6,  1)
test<-t(matrix(famtest,nrow=3))
famp<-pfc(test)
} # }