Thank you for your interest in our work. Please find the Fortran program
and a brief explanatory text attached to this message.
To compile the program, if you are in a unix environment, and have Fortran
77, type "f77 ge_int.f -o ge_int". You can then call the program by
"ge_int". Please report any problems you experience with the program to
me.
For notations and assumptions, please refer to "ge_int.txt", as well as
to Foppa I, Spiegelman D. AJE 1997; 146:596-604.
I also include the non-interactive version of the program ("ge_intn.f").
If you are on unix, the program can be run (after having been compiled as
described above) by
ge_intn < data > out
where data is the name of the input file and out is the output file. For
power calculations, for example, the input file has the form:
p 3
5 .2 2 200 1.5 1.5 2
5 .3 2 200 1.5 1.5 2
5 .4 2 200 1.5 1.5 2
"p" stands for power calculation and 3 is the number of lines (i.e.
calculations). The numbers of the follwoing lines correspond
to:
number of quantiles, gene prevalence,
control-case ratio, number of cases, OR(E|G=0)^(k-1) (top-to-bottom
contrast), OR(G|E=0), (OR(E|G=1)/OR(E|G=0))^(k-1) (to-to-bottom contrast).
If you want to do sample size calculations, use the form
s 3
.9 5 .2 2 1.5 1.5 2
.8 5 .3 2 1.5 1.5 2
.7 5 .4 2 1.5 1.5 2
where "s" refers to sample size calculations, and the following lines
correspond to:
power to be achieved, number of quantiles, gene prevalence,
control-case ratio, OR(E|G=0)^(k-1) (top-to-bottom
contrast), OR(G|E=0), (OR(E|G=1)/OR(E|G=0))^(k-1) (to-to-bottom contrast).
If you wish to receive a compiled version of the program for dos, please
let me know. If you have any further questions, don't hesitate
to contact me.
Sincerely,
Ivo Foppa