From the department of Overthinking-It, I was doing some data manipulations and getting odd behavior from my data smoothing library. Instead of generating interpolated elevation data, the library call was returning NaN (not a number for the uninitiated). After checking the data at various points in the process, I determined that the incorrect outcome resulted from concatenating data sets together. Each data set represents a segment of my cross-continental bicycle ride, so each succeeding segment begins with the ending data of the preceding segment resulting in sequential duplicates when concatenating the data sets. I solved the sequential duplicate data by deleting the duplicate data from the succeeding data set before concatenating.
Problem solved, instead of NaN the interpolation function was returning real values; however, I needed to add the removed data back after processing to output the complete data set. I’m working with arrays, so I initially attempted to figure out how to capture the index of the deleted array items. That gave me a headache, until I realized that one of the variables that I am tracking only changes when a new data set occurs, so I solved the problem by writing out the data and tracking when the variable changed on output and injecting a duplicate of the previous data into the new output thereby restoring my data sets plus the data processing.