|
RadCluster
Graph-based cluster dynamics for irradiated materials
|
Functions | |
| _factor (values) | |
| dict | board () |
| scored (rec) | |
| ref_record (B) | |
| convergence (B) | |
| two_d (B) | |
| main () | |
Variables | |
| HERE = Path(__file__).resolve().parent | |
| DT = HERE.parent | |
| MOD = DT.parent | |
| REPO = MOD.parent | |
| str | SYNC = HERE / ".sync" / "claims" |
| str | OUT = REPO / "docs" / "Formulation" / "verification_campaign" / "figs" |
| str | SCORE_DOSE = "20" |
| LABEL_PT | |
| TICK_PT | |
| LEGEND_PT | |
| list | OBS |
| tuple | CONV_KEYS = ("N_loops_100", "d_100_nm", "N_voids", "d_cavity_nm") |
make_M_figures.py — campaign M: binning convergence, drawn.
Two figures:
M_convergence.pdf the <100> and cavity observables against N_eq, one panel
each (the two <111> panels are omitted), with
the exact arm as a horizontal line. This is the
convergence statement: a rung is converged when its
panel has gone flat AND flat AT the reference line.
Flat alone is self-consistency, which is why the
reference is drawn even when it is the only thing on the
panel that is not a bin-moment solve.
M_2d_sweep.pdf deviation against i_discrete, one curve per I_bin. Kept
as a diagnostic; it is no longer carried in the report,
where the M2D tables make the same point.
Reads the campaign board, so it draws whatever has actually completed and says
so, rather than failing when a rung is missing.
|
protected |
Common power of ten for a set of values, and its label fragment.
Returns (divisor, prefix) so that the panel plots y/divisor and the axis
label reads e.g. "density ($10^{21}$ m$^{-3}$)". Values within two decades
of unity are left alone.
Definition at line 75 of file make_M_figures.py.
| dict make_M_figures.board | ( | ) |
Definition at line 92 of file make_M_figures.py.
| make_M_figures.convergence | ( | B | ) |
Definition at line 118 of file make_M_figures.py.
| make_M_figures.main | ( | ) |
Definition at line 218 of file make_M_figures.py.
| make_M_figures.ref_record | ( | B | ) |
The exact arm, whichever M1 rung is serving as it.
Definition at line 110 of file make_M_figures.py.
| make_M_figures.scored | ( | rec | ) |
Definition at line 103 of file make_M_figures.py.
| make_M_figures.two_d | ( | B | ) |
The 2-D sweep: deviation against i_discrete, one curve per I_bin. This is the figure the M2R ladder could not draw. M2R swept I_bin at fixed i_discrete and found d_111 flat at -5 to -7%; the convergence axis for that observable is i_discrete, so the ladder was moving along a contour of the error surface rather than down it. Plotted this way the curves collapse: the spread BETWEEN I_bin values at one i_discrete is small, while the fall ALONG i_discrete is the whole effect.
Definition at line 169 of file make_M_figures.py.
| tuple make_M_figures.CONV_KEYS = ("N_loops_100", "d_100_nm", "N_voids", "d_cavity_nm") |
Definition at line 72 of file make_M_figures.py.
| make_M_figures.DT = HERE.parent |
Definition at line 37 of file make_M_figures.py.
| make_M_figures.HERE = Path(__file__).resolve().parent |
Definition at line 36 of file make_M_figures.py.
| make_M_figures.LABEL_PT |
Definition at line 50 of file make_M_figures.py.
| make_M_figures.LEGEND_PT |
Definition at line 50 of file make_M_figures.py.
| make_M_figures.MOD = DT.parent |
Definition at line 38 of file make_M_figures.py.
| list make_M_figures.OBS |
Definition at line 62 of file make_M_figures.py.
| str make_M_figures.OUT = REPO / "docs" / "Formulation" / "verification_campaign" / "figs" |
Definition at line 47 of file make_M_figures.py.
| make_M_figures.REPO = MOD.parent |
Definition at line 39 of file make_M_figures.py.
| str make_M_figures.SCORE_DOSE = "20" |
Definition at line 48 of file make_M_figures.py.
| str make_M_figures.SYNC = HERE / ".sync" / "claims" |
Definition at line 46 of file make_M_figures.py.
| make_M_figures.TICK_PT |
Definition at line 50 of file make_M_figures.py.