This folder contains zipped data folders for "The Information-Cost-Reward framework for understanding robot swarm foraging". Each zip file corresponds to one group of data.
The data is used by the analysis python code.

Legend to the folders:
========================

Experiment groups:
0* : parameter optimisation
A* : simulations in static environments
C* : simulations in dynamic environments

Additionally, each folder name also contains the name of the robot controllers that was used to generate data.

Legend to the sub folders:
========================
Each subfolder represents a set of 50 experimental runs from the same task and environment. The subfolders are labelled as follows:

COLLECT_RT1-1* : collection task
PERFORM_RT1-1* : consumption task

*group[N]-[D]* : Heap environments, where N is number of worksites and D is worksite distance from the base middle
*rand[N]-[D]* : Scatter environments, where N is number of worksites and D is worksite distance from the base middle

*_NR[M]* : specifies the number of robots used in the experiment, where M is the number of robots

Additionally, in dynamic environments, each folder is also labelled with the *_DR[T]* parameter, where T is the length of the environment change interval.

Legend to the data files:
========================
Each subfolder contains data files labelled as RunN_global.txt and RunN_infoEvents.txt, where N is the run number.

The "global.txt" files contain a record of the total amount of reward obtained (the "workDone" column) over time. Time is represented by rows.

The "infoEvents.txt" files contain a record of various events that happened during that run. Along with the event type (the "evtType" column), the following data is recorded in the various columns:

robotId: id of the robot responsible for the event. If not applicable, robotId=-1
taskId: worksite id associated with the event
taskVal: reward value per one unit of worksite volume, always 1
taskVol: remaining worksite volume after the event
taskU: "utility" of the worksite, calculated by a robot that recorded the event as (taskVal*taskVol)/(believed distance from the base)
recruiterId: id of the recruiter robot, only relevant to local broadcasters and bee swarm and the ROBOT_RECRUITED and TASK_STARTED events
D: distance of the worksite to the base edge