Future considerations and impovements

[jaytlennon]

I just dug into the Growth_Curves repository for the first time. It's come a long way since I first tried to code this up in SAS. It's much improved and I'm grateful for all the work that people have put into this, especially @meganllarsen @mmuscarella, along with recent contributions from @VKUO2

There are a few things that I found challenging though. If we ever time to make some improvements, I think it would make this code and approach much more user friendly. I should note that I was using the "test.example.R" file, which was written for analyzing data not collected from the plate reader.

1. It seems like it would be nice if the program could run data from a dataframe as opposed to only a .csv file. For whatever reason, this was a little annoying and resulted in some lost time. For example, if you want to remove a wonky row of data, etc.

2. The internal file structure of the package was a bit cumbersome. I would have liked to have called some functions and run the `growth.modGomp" in another .Rmd file. But there seemed to be some peculiarities with where files were located, including bin and output.

3. Perhaps this has more to do with the math and data, but I found that some of the parameter estimates were very sensitive, maybe do to starting conditions (intercepts and A). @mmuscarella suggested this could be do to maximum likelihood estimation and getting stuck in parameter space.

I'm sure I could play around with these, and maybe I will, but I also thought I would share some thoughts for the next user while things are fresh in my head.

[mmuscarella]

Made a minor change to allow users to define where output data and temporary plots would be saved. Reminder, you only need the four files in the bin directory. These files should be put in a bin directory in the repo. This directory should be in the top directory of the repo (as with our other projects and as is, to my understanding, standard protocol).