I want to create a matrix that stores the sensitivity and specificity of my probit model at all thresholds from .01 to 1.00. So I create a 100x3 matrix and wrote this loop:
Code: Select all
%dep = "my_binary_var"
%eq = "my_equation"
matrix(100,3) my_matrix
for !i = 1 to 100
!threshold = !i/100
freeze(iteration_tab) my_equation.predict(!threshold)
my_matrix(!i,1) = @val(iteration_tab(13,2))
my_matrix(!i,2) = @val(iteration_tab(13,3))
my_matrix(!i,3) = !threshold
d iteration_tab
next
The thing is it seems a little clunky and unsophisticated because I have to freeze and delete the prediction table each time. Is there a better way I can approach this task?
