Hi,
I was wondering if there is an option in the csipolate command or any other command which allows to perform cubic spline interpolations with a monotonic side condition. I am using cubic interpolations to fill in missing values for average tax rates, of which I only have known values for incomes of (in thousand): 4,6,8,10,12, 15, 20, 30, 40, 50, 60, 80, 100, 200.
As the average tax rates of 100 and 200 are very close (20.47 and 20.51), Stata generates values between 100 and 200 that are bigger than 20.51 (inverted u-shape), which of course makes sense for a cubic interpolation but is not possible in the real world, as the tax function should increase monotonically.
Many thanks,
Linus
I was wondering if there is an option in the csipolate command or any other command which allows to perform cubic spline interpolations with a monotonic side condition. I am using cubic interpolations to fill in missing values for average tax rates, of which I only have known values for incomes of (in thousand): 4,6,8,10,12, 15, 20, 30, 40, 50, 60, 80, 100, 200.
As the average tax rates of 100 and 200 are very close (20.47 and 20.51), Stata generates values between 100 and 200 that are bigger than 20.51 (inverted u-shape), which of course makes sense for a cubic interpolation but is not possible in the real world, as the tax function should increase monotonically.
Many thanks,
Linus
Comment