I would like some thoughts on a way of quantifying a simple measure of lipid chain order parameter from a gromacs file. The g_order will give me an average and an error term for each of the specified atoms in a lipid chain. However, I would like a single quantitative value that represents the order of the whole lipid; a very ordered lipid would return a value of 1, whilst a disordered lipid would return a value close to 0. These are just arbitrary numbers I am using, but it kind of highlights what I am after. For example, if I take a a simulation of a lipid bilayer, 30 nm by 30 nm, and divide it into 3 nm by 3nm segments, I would like a single measure to suggest how ordered the lipids are within each segment. You could imagine how I could then use this as a function of distance from a point such as a protein. I would rather not have the average order per atom, as this would leave me with many plots.