Data file

Previous: Overview of the programs Up: Overview of the programs Next: Program options

Data file

The data file produced by the simulation consists of a series of records, with a field for each particle. The number of particles in each record is not assumed to be constant, although it is always constant in this simulation.

The data itself is stored in native format. On the Sun Sparc CPU used, there are 64 bit IEEE format double precision floats. A problem with using binary files is that the data becomes machine type dependent. This means a conversion process is required if one wants to transport a data file to another type of machine.

However, there is a cost to storing the data in ASCII format. Both in file size, and it reading and writing cost. Considering that the data files easily grow to 20Mb in size, fast post-processing is desired, and actually visual inspection of the data is of limited interest, the portability benefits of an ASCII file were of less importance. It was also assumed that the initial conditions, and/or a random seed would be sufficient to reproduce the data in a reasonable time given the simulation program.

A library file was written to read the data file back into memory.

mcr@ccs.carleton.ca