TA3 Age Estimation method: Shiny app replicate

This Shiny app is based on the TA3 Age Estimation software (beta version) created by Ron Richardson and Stephen Ousley which can be found here. Here is some information to contextualize the TA3 method:

• It was made by Biological Anthropologists to estimate the age-at-death of unidentified human remains.
• The TA in TA3 stands for Transition Analysis, a method originally based on logistic regression and Bayesian probabilities.
• The new TA3 method uses machine learning, specifically a random GLM model to estimate age-at-death from skeletal traits using data from other individuals with known-age.

The original code from Ron Richardson's GitHub page was modified and used in the creation of this Shiny app.

Because there are 121 skeletal traits that can be evaluated for the TA3 method, I wanted to find a way to set these many radioButtons in the Shiny UI without having to manually set them. For this, I used a lapply loop that depended on information from a CSV file. Here is the head of that CSV file:

           TraitDBName                  TraitText   0    1 2 3 num cols
1 parietal_depressionL Parietal depression (left) abs pres       1    4

The next challenge was retrieving those radioButton input values to be able to use them in the random GLM model. This was also done using a lapply loop.

    fieldsAll <- c("IDInput", "indDegInput", "recorderInput", "obsDateInput", "noteInput", 
                   lapply(1:121, function(i) {
                     paste0("radioInput", i)  }))
    list_res <- lapply(fieldsAll, function(x) input[[x]])

I myself am a Biological Anthropologist very interested on what the TA3 method can achieve. I set out to replicate what the TA3 software does in a quest to understand how the method works.
This is just the beginning with this Shiny app as I would like to add the following features to it:

• Allow the user to submit more than one case and then estimate their age using a loop.
• Allow the user to input multiple cases by uploading a CSV file with a specific format.
• Improve the method by evaluating other machine learning algorithms!

Here is the code for the Shiny app. Here is my app.