Author: ethanc751

Today was the last day of the Work Experience Program. Shortly after noon, Emily and I met again to combine what work we had done, and after some struggling, we managed to succeed and present the finished product to Dr. Aziz and Leah.

Our hope is that they will be able to use our version of the model, along with instructions Emily wrote, as an activity for their students. The main thing about this experience which sets it apart from others was how hands-off it was. We were mainly free to go our own way and learn at our own pace, with minimum input and flexible schedules..

Today, I finished what I started yesterday in separating the lignin value from the fungus and putting it in the background. Now, the fungus spreads evenly across the simulation as it chases the glucose produced from the enzymes breaking down the lignin, which fits reality much more closely in how it creeps across the wood in the background, much like real mold.

Because lignin is now tied to the background, it means more is readily available, so more of it gets broken down into glucose, which is then used by the fungus to stimulate growth, leading to a much higher amount, about 8 times as much. I expect this is how it is in reality as well, as the cells should have a fairly high density. With this crucial step, the final model nears completion. Tomorrow, before the last meeting, Emily and I will combine codes one last time.

Dr. Aziz told us today that what we have completed so far is similar to what he would expect a graduate student to do, and he said that we would be receiving high marks if we were his students.

Today was supposed to be mainly debugging and wrapping up, but my model hit a crucial wall because I wanted the fungus in the model to be sustainable and to make it grow endlessly, or at least like Dr. Aziz said where its food source, wood chips, wouldn’t have to be replaced for a considerable amount of time. However, the lignin in my model was running out incredibly quickly. I tried fixing it, but all this resulted in was an ever-expanding fungus population as the fungus didn’t die off before the lignin was replenished, and it was revitalized each time, shown here.

The aging fungus, however, didn’t produce the enzymes used to break down the pollutants, so even though there was a lot of it, most of it was useless. It also caused copious amount of computer lag due to the incredibly high numbers of simulated fungus later on. The problem was that the lignin amount was tied to the fungus, not the background like it was supposed to, but separating them will take much work. I’m attempting to do that now, but I don’t know if I’ll have enough time in the WEP to do so.

Today, we were left to our own devices as we weren’t required to meet with Dr. Aziz. I continued my work on simulation water flow through a fungal filter by making nutrients, O2, Co2, etc. spawn at the bottom of the screen and move upwards, vanishing as it reaches the top. However, when the speed of the water is set above 1, nothing moves upwards properly, leading to a giant dead zone where only the top and bottom rows have these nutrients.

The problem is this one line of code, which is supposed to make the nutrients disappear if they’re about to exit the top of the screen, but is apparently going gung-ho on everything and deleting it all as soon as it spawns in.

As I don’t know how to solve the issue, I’ll have to consult with Dr. Aziz tomorrow.

This morning, Emily and I worked together in a call to combine aspects from both of our models into a single new, seizure-inducing model.

Compared to the original base model we were given, Emily’s model had large pieces of new code that could be easily copied, as opposed to mine which had smaller bits scattered throughout, so we decided to copy her new work into mine. She did have an issue in which the O2 levels would go negative for no apparent reason, but working together, we identified and fixed the issue, which was a small piece of code that had been inserted in the wrong place.

As Emily is now working on separating the lignin from the fungus and putting it in the patches instead, I am now currently working on making the model simulate the flow of water through a filter, and I look forward to Emily completing her work so I may incorporate it as well.

After this morning’s meeting, most of today was spent fine-tuning the model. The one new feature I did add was making it so the glucose would slowly spread out over a large area, causing the fungus to have to chase after it more. Most of the variance for the temperature setting was created by using a slightly-controlled random number generator, which made the increase and decrease in effects very linear, but I received an equation from Emily which would make it a more natural bell curve, shown below.

This marks the first step in combining our models. Tomorrow, we will meet again and work to combine my fungus growth and temperature control with her dissolved o2 and PH controls, among other things, into one complete model, and we will go from there. I was planning to make all of the simulated organisms and nutrients move from one side of the screen to another to simulate water flow, but I questioned Dr. Aziz and he said focusing on the interactions in the existing model was more important, so I’ll scrap the idea for now.

Once again today, we met with Dr. Aziz and Leah to review the work we had done last time. This time, we went over individually what we could do to improve our models and went over several mathematical formulas that could improve the realism of our work. I was suggested that I could implement an aging system to the simulated fungus in which the enzymes they produce decreases the longer they’re alive, which I added. I also tried swapping around the values in the equations for the temperature slider I added last time. I suck at math, so it took me an hour to figure out, but I eventually got it. I also tried to make it so the simulated glucose is able to randomly move around the model, but after much experimenting, I was unable to figure it out. I eventually just turned to googling the answer and was able to piece together the process, shown below, from several different pages. In each patch, or in each small part of the simulation, it causes a random amount of glucose to move to a random nearby patch each time the simulation progresses. We were told by Dr. Aziz that next week, we will be working mainly independently, and he will only check in with us once or twice.

Today was another day of experimenting with the modeling program to prepare for our independent work. This morning, we met with Dr. Aziz and Leah again to overview what we did last time. Afterwards, we played with the existing model that Dr. Aziz had been using to model competition for nutrients between the white rot fungus and any bacteria that may come through the same location.

The model uses agents that follow programmed codes in a simulated environment, with various sliders and buttons that the user can employ to change various aspects of the simulation, such as the number of nutrients or agents involved. To begin our coding, we were given a list of suggestions we could try implementing to make the model more realistic, such as including the temperature or PH of the water.

Below is an image of the model.