Create means of applying corrections to data and creating compound timeseries

[gdelaplante]

AquaCache is not currently set up to accept or apply corrections to data. In the even that this is needed/desired, a process similar to that used by Aquarius can be considered. A table structure has been created to hold this information and is in the dev/data_corrections.R file.

Missing is a views table that mirrors measurements_xxxx tables but applies corrections to the data passed through. Notes to this effect are included in the file.

[gdelaplante]

Table structure and SQL operations created: public.corrections holds the corrections and their periods of applicability public.correction_types holds information regarding the corrections, their priority of application, etc. function apply_corrections contains the logic for applying corrections view public.measurements_continuous_corrected and public.measurements_discrete_corrected apply this function to output corrected timeseries

Still missing:

1. Determining what to do with the calculated_daily table. Currently this takes values from table measurements_continuous, but should this be switched to the corrected table at the expense of a bit of speed?
2. Creating a GUI means of applying corrections.

An additional useful functionality is creating calculated (such as specific conductivity), compound (such as multiple climate stations), or derived timeseries. A draft workflow is in file dev/calculated_derived_tsids.R. Implementing this would require:

• A new 'timeseries' table (say, timeseries_all?) that holds all raw AND calculated/composite timeseries
• New tables for compound timeseries to hold things like which timeseries to use and for which time periods, equations to apply, etc.
• Adjustment of the corrections workflows to work on timeseries from the 'timeseries_all' table such that the measurements_xxx_corrected view tables output corrected data from raw AND calculated/compound timeseries.