12.5.2. Data filtering
To extract the relevant data, the files first need to be filtered. A custom-made filter algorithm erases rapid-succession-readings of the same transponder-ID at antennas of the same scanner, which occur when a tagged bee lingers beneath the scanners for too long e.g. by being blocked by a flow of opposing traffic. Another custom-made filter programmed to search only for times that take longer than 30 seconds helps to erase many short-duration stays outside the hive, which possibly occur during guard duties and quick movements into the hive or out of the hive. To calculate the time needed from hive to feeder and back again, scanner data at the hive entrance and at the feeder need to be synchronized. Three different algorithms are required:
- For calculation of foraging trip duration and flight times back and forth.
- For feeder stay duration.
- For in-hive
It is advisable to introduce additional variables which include entry numbering to avoid errors during data resorting in same-second entries. If this is not done, same-second entries might be inverted by the program, leading to errors in data evaluation.