Input formats¶

kippy chooses a reader from the supplied path. The checks are ordered: a basename beginning with seq. or ending with .seq selects MONASH, a path ending with .kipp selects the raw stream reader, and a path for which <path>/profiles.index exists selects MESA. Every other path is passed to the KEPLER convection reader; the .cnv suffix itself is not checked.

KEPLER .cnv¶

The KEPLER reader consumes the convection zones, mass and radius coordinates, and energy layers stored directly in the sequential unformatted .cnv records. These native energy records provide the nuclear and neutrino bands used by the plot. The reader does not register generic source columns as selectable colour fields.

The Fortran reader supports record versions 10600 through 10699 and record versions 10700 and later. Versions below 10600, including 10400 and 10500, are rejected with an error.

MESA profiles directory¶

A MESA input is a directory containing profiles.index and the referenced profileN.data files. profiles.index maps model numbers to profile numbers; each selected profile supplies the age and mass-radius grid for one plotted model. Convection zones come from mixing_type, the epsnuc layer comes from eps_nuc, and eps_nuc_neu_total supplies a separate neutrino layer.

MESA profiles may be sparse in time rather than present at every model. The band tracer bridges consecutive loaded profiles into continuous polygons.

Kippy offers a curated set of MESA columns as colour fields: logT, logRho, logL, luminosity, logP, pressure, opacity, entropy, and velocity. Only columns present in the run are registered.

Raw .kipp binary stream¶

A .kipp file is a custom file contiguous little-endian float64 cell dump. Cells are grouped by monotonically increasing model number, with one row per zone. The sidecar header defines the column count and column names, including the model, age, mass, radius, mixing, and single eps columns required to build the plot. The mixing column supplies convection, while the eps column supplies one net energy layer; no separate neutrino layer is created.

For a data path such as path/to/file.kipp, Kippy first looks for path/to/file.kipp.hdr beside it. If that file is absent, it looks for file.kipp.hdr in the working directory.

The reader streams the data twice. The first pass builds coordinates, convection, and energy while finding the run-wide range of each generic colour field. The second pass quantizes those fields. This keeps the raw columns for the full run out of memory. Every named column not used for model metadata, coordinates, mixing, or eps is available as a colour field.

MONASH seq¶

The MONASH reader expects a gfortran sequential-unformatted file with one record per model. Mass coordinates are reconstructed from omx as mass * (1 - omx)^3; the cells are already ordered from centre to surface. Convection zones come from kcvtn.

The epsnuc layer is approximated from the luminosity change per mass shell, dL/dm. This is not a pure nuclear energy-generation rate because the luminosity difference also includes gravothermal terms.

The registered colour fields are Temperature, Density, Pressure, Luminosity, H_reaction_rate, He3_reaction_rate, He4_reaction_rate, C_reaction_rate, N_reaction_rate, O_reaction_rate, and Other_reaction_rate. These names are case-sensitive. Like the raw stream reader, the MONASH reader uses one pass to find field ranges and a second pass to build the quantized field layers.

Colour fields¶

Use color <column> to select a registered non-structural column. The raw .kipp reader registers every named non-structural source column; the MESA and MONASH readers register the curated fields listed above. Use fields to list the selectors and ranges available for the loaded input.

Field values are quantized into 24 contour bins over the minimum and maximum across the loaded run. Scaling switches automatically from linear to logarithmic when the complete range is positive and spans more than two decades. These generic fields are separate from the built-in epsnuc energy overlay.