Output directory

Each run writes a self-describing directory under radcluster_code/output:

output/YYYYMMDD_HHMMSS_<git-hash>/
    provenance.md
    results_t.npy
    results_y.npy
    summary.csv
    diagnostics.txt
    plots/

where

• provenance.md contains all input tables with the overrides applied, the user selections, the solver configuration, the run time, the machine and the CVODE statistics
• results_t.npy contains the time points
• results_y.npy contains the full ODE solution. See State layout for the meaning of its columns
• summary.csv contains a one-line summary of the end of the run
• diagnostics.txt contains the conservation residuals and the end-state diagnostics
• plots contains the figures in PNG format, and the file plot_data.pkl from which the figures can be drawn again without a new run

The summary file

The file summary.csv holds one row with the columns:

solver physics_option T_K G_dpa_s t_end_s dose_dpa C_SIA_tot_m3 C_VAC_tot_m3 C_He_tot_m3 mean_n_i mean_n_v swelling delta_FP delta_He

Because every run writes one row, the summaries of a campaign concatenate directly into one table. The row is built by the function summary_csv_row() in post_process.py.

Provenance

The git hash and the complete resolved parameter set are written with every run, so that a figure can be traced back to the code and the inputs that produced it. Runs never overwrite one another.

Note that the folders build and output are not tracked by git, with one exception: the designated reference run is tracked, because the verification campaign reads it by name.

Size distributions

Size distributions are drawn as per-bin densities \(dc/dn\) or \(dc/dD\), as stair segments whose integral is the population in the displayed range. This is the quantity that compares with distributions measured by transmission electron microscopy.

See also

Input workbook, Verification and diagnostics