Our real world scenario involves R scripts that process raw smoke monitoring data that is updated hourly. The raw data comes from various different instruments, set up by different agencies and transmitted over at least two satellites before eventually arriving on our computers.
Data can be missing, delayed or corrupted for a variety of reasons before it gets to us. And then our R scripts perform QC based on various columns available in the raw (aka “engineering level”) data.
Logging is one of the differences between toy code (even very useful toy code) and production-quality code. Read on for an easy way to do this in R.