rShiny, Different values from the predFit function when nlsLM is in a function or not

Question

I\'m trying to make a regression line from a data table and since I have to use the nlsLM function in many places I thought I would make it into a function instead. The prob