Here will be an explanation of how to run the script to QC and the debug script in case of problems

Run Scripts

There is two scripts in the main folder of the project (/sapfluxnet):

  1. main_script.R

  2. debug_script.R

main_script.R

This script is in charge of loading the sapfluxnetQC1 package, move any new data from received to accepted folder, and start the QC process.
It is a very small and simple script, which takes advantage of all the functions present in the package. It is intended to be executed by a cron job (system level) or by the button present in the SapFluxNet Monitor app (interactive level). It also can be ran in a RStudio Server session (high interactive level).

debug_script.R

This script is intended to be executed manually only when an error during the QC process is found. It replicates all the steps made in the QC process, but outside of the Rmd report. In this way, all the intermediate objects created by the QC process are available and can be studied to identify the error source/step.
In this case, error means that something went wrong in the QC resulting in the unavailability of the report or really weird reports. This can lead to find errors in the functions (data kinds we didn't account for before) or errors in the data (not in the package). Depending on the error found, solution can involve improving troublesome function/s, ask contributor politely to amend data or make easy and small manual changes in the data.

RData objects

Qc_report.Rmd template also save the intermediate objects created during the QC process in a RData file in the data folder.
This file can be opened in a RStudio Server instance, allowing access to the objects without running the debug script. This can be useful in very large datasets, where running the script could be very time consuming.



sapfluxnet/sapfluxnetQC1 documentation built on May 29, 2019, 1:50 p.m.