Output Files

Wile MABE runs it generates data related to each organism. This data is periodically saved by an archivist in csv format output files. In addition, depending on the modules and setting in use, non-archivist data files may be saved. The names of files saved by archivists follow certain rules - data files contain data collected from evaluations and organism files contain data needed to reconstruct organisms (usually genomes, but also non-genome based brains, etc.). Generally, data files contain time series data (i.e. each line in the file correlates to a particular time or update). Files with snapshot in their name will also contain a number. Snapshot files contain data for a collection of organisms which were alive at the indicated update. Lastly, LOD in a file indicates that the file contains data for organisms on "line of descent".

Files generated by the Default Archivist:

pop.csv contains population averages for a subset of data at various time points. The data saved to the pop.csv file are automatically determined by MABE (each component of MABE tells the Archivist what data they think is interesting). Each line of this file represents one time point (update). This file provides a curated view of the evolving population over time.

max.csv contains data related to the single organism in the population with the highest "score" (as determined by the optimizer) at various time points. Each line of this file represents one time point (update). This file lets you see, in detail, how the best organisms are doing over time.

Each snapshot_data_#.csv stores a snapshot of the state of the population at update #. Each line in these files contains information about one organism in the population.

Each snapshot_organisms_#.csv contains a snapshot of organisms in the population at update #. Each line in these files contains organism data for one organism in the population. This file can be used to reconstruct an entire population.

Files are generated by the Line of decent with aggressive Pruning(LODwAP) Archivist.

LOD_data.csv is similar to max.csv but contains data related to organisms on the line of descent of an asexual population. Each line of this file represents one time point (update).

LOD_organism.csv contain organism data on Line of descent and line the LOD_data.csv file, each line of this file represents one time point (update)..

Files are generated by the SnapShot with Delay (SSwD) Archivist.

Each snapshot_data_#.csv stores the data for organisms that were alive at # and still have surviving offspring after some delay. Each line in these files contains information about one organism.

Each snapshot_organisms_#.csv contains organism data for organisms that were alive at # and still have surviving offspring after some delay. Each line in these files contains information about one organism.

snapshot_data and snapshot_organism files must be requested.

snapshot files (particularly organism files) can be large so these files are only produced if you have switched on writeSnapshotDataFiles and/or writeSnapshotGenomeFiles in the Default and/or SSwD archivist.

Line of Decent vs. Snapshot and Snapshot with Delay files

the Line of Decent with Aggressive Pruning (LODwAP) Archivist produces file with this format:

 LOD_data.csv   LOD_organisms.csv

In these files, each line contains information about the organism on the line of decent (i.e. each line in the file is a diffrent point in time). LODwAP assumes asexual reproductino and that speciation is not occurring (for more on this, see: archivists).

Snapshot and Snapshot with Delay (SSwD) files have this format:

snapshot_data_0.csv       snapshot_data_30.csv     snapshot_organisms_20.csv
snapshot_data_10.csv      snapshot_organisms_0.csv    snapshot_organisms_30.csv
snapshot_data_20.csv      snapshot_organisms_10.csv

the number in the name of each of these files tells you the update that the data in that file relates to. (i.e. genome_20.csv contains genomes from organisms alive at update 20.)

Other Output Files

Various modules in MABE output their own files. These files are managed by modules which create them. Here are some examples:

• Wire Brains can record the state of a wire brain which displays how current flows thought the brain
• Berry World can record the state of the world which can help define the strategy organisms are using.