Parse FLUKA Output

Nucleide reads FLUKA USRBIN mesh tallies from .lis output files without running FLUKA itself. This tutorial shows the Python API; equivalent Rust types live in the nucleide-fluka-io crate.

USRBIN tallies

read_usrbin parses every USRBIN block in a .lis file into one UsrbinTally each: detector metadata, mesh bounds, track-length data, and percentage errors as plain lists — no meshing layer.

from nucleide.fluka import read_usrbin

tallies = read_usrbin("fixtures/fluka/fluka_usrbin_single.lis")
t = tallies[0]
print(t.name, t.particle)

nx, ny, nz = t.dims()
print(nx * ny * nz == len(t.data) == len(t.error))
print(t.x_bounds)  # nx + 1 bin edges

data and error are flat in the file’s A(ix,iy,iz) order (first index fastest), so len(data) == nx * ny * nz and each of x_bounds, y_bounds, z_bounds carries one more edge than bins.

Multiple detectors per file

A .lis file can hold several USRBIN blocks; each becomes one tally. The sample with degenerate binning shows that per-axis bin counts need not be ordered largest-first:

tallies = read_usrbin("fixtures/fluka/fluka_usrbin_multiple.lis")
print([t.name for t in tallies])

Sample files

Sample reference files live under fixtures/fluka/ (single, multiple, and degenerate binning cases plus the generating input deck). The test suite checks that Nucleide reproduces them byte-for-byte where exact reproduction is intended; see the sample data files index for the full list.

See the Python API reference for the full signatures.