What is the "Time" column in the fix ave/correlate/long output?

I am using the following code to compute the components needed to calculate the stress relaxation modulus G(t) according to Eq. 4 in Ref:

compute c_stressA all stress/atom

compute pA1 all reduce sum c_stressA[1]
compute pA2 all reduce sum c_stressA[2]
compute pA3 all reduce sum c_stressA[3]
compute pA4 all reduce sum c_stressA[4]
compute pA5 all reduce sum c_stressA[5]
compute pA6 all reduce sum c_stressA[6]

variable nxy equal c_pA1-c_pA2
variable nxz equal c_pA1-c_pA3
variable nyz equal c_pA2-c_pA3

fix 5 all ave/correlate/long 1 1000 c_pA1 c_pA2 c_pA3 c_pA4 c_pA5 c_pA6 v_nxy v_nxz v_nyz type auto nlen 16 ncount 2 ncorr 30 file output-folder/output-file.out.correlate.txt

Naïvely, I would expect the output to have rows, one row for each time-step at which to output, containing the auto-correlated values of all the variables & computes I pass.
However, the output of the fix ave/correlate/long contains sections, one for each time-step at which LAMMPS outputs, each section with an additional column “Time”.

My question is, what does this “Time” column represent, where does it come from, which “Time” do I want to use to plot G(t) vs t?

This is a question I should in theory be able to answer by reading the code of LAMMPS, which I am currently doing, but it seems to be more involved than I might have hoped. I would update this post if/when I do find an answer.

My current understanding is, that the “Time” column in the output of the fix ave/correlate/long command in LAMMPS represents the time delay between two data points being correlated. It is calculated by multiplying the number of time steps between two data points by the time step size.
Therefore, if I need <\sigma_xy(0) \sigma_xy(t)>, what I want is only the last line of each section.

Hi @GenieTim,

As far as I can remember the output of ave/correlate/long is the same format as the output from ave/correlate. According to the doc:

  • The first column has the time \Delta{}t (in time steps) between the pairs of input values used to calculate the correlation, as described above.