Team:Wageningen UR/Notebook/beehave
Beehave
July
Week 4
Orientation literature research, found articles on BEEHAVE model. Proceeded to read the modeling articles BEEHAVE is based on. Also installed NetLogo and ran a few standard simulations.
August
Week 1
Started looking into BEEHAVE model and figuring out how to integrate BeeT into it. Identified ways of adding mite mortality to the model. Made a very rough first approximation of mite mortality and fixed resultant bugs.
Week 2
Start looking into how to have transport BeeT into the hive. Initial idea is to add an additional 'BeeT' patch, turns out patches are hard coded into the model and adding in a new one is difficult. Got started with a new BeeT module 'BeeTproc' since the changing the patch was a dead end.
Week 3
Started using a local github depository since changes were getting difficult to keep track of. Added comments to all the code I added and put everything into github. Also coupled BeeT in the hive to mite mortality. Added period at which BeeT patch is present.
Week 4
Worked on getting initial results by increasing year round mite mortality, worked on understanding RNetLogo and getting some statistics for results. More background literature research as many aspects of system still unclear
September
Week 1
Had a call with a beekeeper (Johan Calis) and based on that call found articles regarding sugar water transport in the hive. Started working on alternative application method using bee bread.
Week 2
Added in degradation for BeeT in the patch and the hive. Started adding in code to handle bee bread. Unclear how to determine how much ends up at the larvae.
Week 3
Realised that consumption of pollen and sugar water is included in BEEHAVE. Proceed to couple BeeT transport in the hive to consumption patterns of bees and larvae. Fixing more bugs.
Week 4
Finish up final version of BeeT module, include reporters to make sure everything works correctly. General cleanup of code. Get access to server and start figuring out how parallel works.
October
Week 1
Find bugs in my RNetLogo code related to the server. Waste time tracking them down and fixing them. Mess up parallel so that it repeats the same code 20 times instead of breaking it up into 20 pieces. Improving R code so I can save results while parallel is running. Start running code on server.
Week 2
Start writing wiki related material. Get results from the server and start analyses. Realise my server data was not correct, rerun on the server.
Week 3
Run more in depth runs to figure out what is going on exactly. In results, we ran the simulation for figure 20B with the scaling variable for mite mortality of 2, degradation outside the hive of 0.004 and degradation inside the hive of 0.006. For figure 20C, the scaling variable for mite mortality is 6, degradation outside the hive 0.004 and degradation inside the hive 0.006. The same paramater set was used in figure 20D, only now with a starting concentration of 10,000 mites.