We have collaborated with Vilnius-Lithuania. They have given us feedback on our code, helped with camera settings in the Blender Software and we have given them user experience feedback on their AR project.
Removed global variables in the analysis script, this is considered a bad practice and can cause problems that are hard to track down. Further comments on the optimizers and narrowing down their use has been implemented. Lastly more comments on the accuracy function and comments on what is a function's input and or output. In the model code we changed parameter descriptions and a significant alteration to comments on functions, detailing input and outputs.
A special thanks to Emilija from Vilnius-Lithuania that demonstrated tips and tricks on how to better use the camera. This allowed us to create better videos that allowed us to show different aspects of the data generated by our model.
We also had a collaboration with GunnVistaPingry_US, a high school team in the United States. They gave a very good review of our code and as a consequence we included the capability of saving the neural net created by the iGEMDataAnalysis code, in addition we document the analysis code more on the readme. The additional documentation was added to make it clear how to change class in parametersforModel.txt. We also uploaded sample CSV data, so the end user could classify / visualize without running the model. They also gave us very good feedback that we would want to implement in the future, more try-catch to handle expectations, close files when we don't use them and create a dictionary from the parametersforModel.txt instead of reiterating many times.