Standard log file format¶

COMPAS can produce log files in several formats:

Hierarchical Data Format version 5 (HDF5)[1]

Comma Separated Values (CSV)

Tab Separated Values (TSV)

Plain text: space separated values (TXT)

The log file type is set using the --logfile-type program option.

Standard CSV, TSV, and TXT log files are human-readable files, and formatted in a similar fashion. Each standard CSV, TSV, and TXT log file consists of three header records followed by data records. Header records and data records are delimiter separated fields, and the fields as specified by the log file record specifier.

The header records for all standard CSV, TSV, and TXT log files are:

Header record 1: Column Data Type Names
Header record 2: Column Units (where applicable)
Header record 3: Column Headings

Column Data Type Names are taken from the set { BOOL, INT, FLOAT, STRING }, where

BOOL

indicates the data value will be a boolean value.

Boolean data values will be recorded in the log file in either numerical format (1 or 0, where 1 = TRUE and 0 = FALSE), or string format ("TRUE" or "FALSE"), depending upon the value of the --print-bool-as-string program option.

INT

indicates the data value will be an integer number.

FLOAT

indicates the data value will be a floating-point number.

STRING

indicates the data value will be a text string.

Column Units is a string indicating the units of the corresponding data values (e.g. "Msol*AU²*yr^-1", "Msol", "AU", etc.). The Column Units value may be blank where units are not applicable, or may be one of:

Count

indicates the data value is the total of a counted entity.

State

indicates the data value describes a state (e.g. "Unbound" state is "TRUE" or "FALSE").

Event

the data value describes an event status (e.g. "Simultaneous_RLOF" is "TRUE").

Column Headings are string labels that describe the corresponding data values. The heading strings for stellar properties of constituent stars of a binary will have appropriate identifiers appended. That is, heading strings for:

STAR_1 PROPERTY::properties

will have "(1)" appended

STAR_2 PROPERTY::properties

will have "(2)" appended

SUPERNOVA_PROPERTY::properties

will have "(SN)" appended: any column with a header with a suffix of "(SN)" represents an attribute of the star undergoing a supernova event, either just before the supernova [e.g., Mass_Total@CO(SN)] or just after the supernovae [e.g., Mass(SN)].

COMPANION_PROPERTY::properties

will have "(CP)" appended: any column with a header with a suffix of "(CP)" represents an attribute of the the companion after the supernova event.

HDF5 files are not human-readable. The HDF5 file format supports large, complex, heterogeneous data, enabling the data to be stored in a structured way in a single file. When the HDF5 format is specified for COMPAS log files, a single HDF5 file is produced for non-detailed output log files, containing all non-detailed output log files described above. Detailed output files are created, as for other logfile types, as individual files (in this case, HDF5 files), in the ’Detailed_Output’ container directory.

Each file described above is created as a group within the HDF5 file, with the name of the group set to the name of the file (e.g. "BSE_System_Parameters"). Each column in the files described above is created as a dataset within its corresponding group in the HDF5 file, with the name of the dataset set to the column header as described above (e.g. "Mass(1)"). Each dataset in an HDF5 file is typed, and the dataset data types are set to the column data types as described above. The column units described above are attached to their corresponding datasets in the HDF5 file as attributes.

Footnotes

BOOL	indicates the data value will be a boolean value.
	Boolean data values will be recorded in the log file in either numerical format (1 or 0, where 1 = TRUE and 0 = FALSE), or string format ("TRUE" or "FALSE"), depending upon the value of the `--print-bool-as-string` program option.
INT	indicates the data value will be an integer number.
FLOAT	indicates the data value will be a floating-point number.
STRING	indicates the data value will be a text string.

Count	indicates the data value is the total of a counted entity.
State	indicates the data value describes a state (e.g. "Unbound" state is "TRUE" or "FALSE").
Event	the data value describes an event status (e.g. "Simultaneous_RLOF" is "TRUE").

STAR_1 PROPERTY::properties	will have "(1)" appended
STAR_2 PROPERTY::properties	will have "(2)" appended
SUPERNOVA_PROPERTY::properties	will have "(SN)" appended: any column with a header with a suffix of "(SN)" represents an attribute of the star undergoing a supernova event, either just before the supernova [e.g., Mass_Total@CO(SN)] or just after the supernovae [e.g., Mass(SN)].
COMPANION_PROPERTY::properties	will have "(CP)" appended: any column with a header with a suffix of "(CP)" represents an attribute of the the companion after the supernova event.