A simple cuda-gpu simulation of random walk in a crowded geometry. The repository contains metadata used by NSight so it can be easily set up in this IDE.
./obstructed_tracer [mode] (mode specific arguments)
Mode can be one of:
- perform_walk,
- analyze.
(mode specific arguments): [input file] [output files prefix]
It reads the parameters from input file and performs random walk. The output mean square displacement data and, if desired, trajectories data will be saved as
[output files prefix]_msd.tx
and
[output files prefix]_[simulation index]_[trajectory_index].txt
(mode specific arguments): [input file] [msd file]
It reads the parameters from [input file] and calculates the diffusion coefficient D and exponent α for last two orders of <r2>(t) and <var(x)+var(y)>(t) from [msd file]. It also computes the correlation of x and y for the last point and for the middle one on the log scale. It assumes that [msd file] file was generated using the same input file as given.