This section describes the format which data collected from a simulation are saved in.
The CreativeAnalyzer class records data for a single simulation. The AnalyzerAggregator class combines data from multiple simulations.
When a simulation is run, a directory is created where data about that simulation is saved. This directory will contain a subdirectory named analyzer, where simulation data is saved.
An analyzer directory contains used.cfg, a copy of the simulation configuration file, as well as files for each metric which was recorded.
Metric files are named after the metric they contain, such as complexity_avg or genome_length_avg. These files contain numpy arrays in text format. Generally, each row corresponds to a generation, with generation 0 at the top of the file, and generation 1000 at the bottom.
The gens_until_x metrics are slightly different. Instead of generation number, each row corresponds to a value from the thresholds list in the threshold-metrics section of a simulation configuration file.
When a comparison is run, multiple analyzer directories are aggregated into a single comparative-analyzer directory. Again, each file holds a numpy array in text format which corresponds to a metric. Generally, each column of a metric file corresponds to a type of simulation which was run for that comparison.
Some metrics such as the gens_until_x metrics also contain data on the standard deviation and number of blank values, named gens_until_x.stddev and gens_until_x.blanks respectively. During graphing, blanks are replaced with generation_limit.