AutoDropOutliers: Checks for outliers in flux data.
In bmcnellis/sapflux: Analyze Sapflux Traces

Description Usage Arguments Details See Also Examples

Pre-processing sapflux data can be time consuming - while the whole process hasn't been automated, some outliers are easy enough to pull automatically.

1 2	AutoDropOutliers(flux, byIQR = FALSE, drop_time = FALSE, max_time_gap = 30, min_data = 5)

`flux`	A 'flux' class object.
`byIQR`	Logical - use interquartile range to identify outliers?
`drop_time`	Drop empty time rows? TRUE or FALSE
`max_time_gap`	Maximum time length in minute to interpolate.
`min_data`	Minimum amount of data a column needs to be included, expressed as a percentage of total data.

In order, AutoDropOutliers checks the following:

1) Are values above an 'absolute_maximum' defined in package metadata?

2) Are all values in a data column NA?

3) Do values occur before the probe install date, or after it's removal date?

4) Are there missing values in < 30 minute gaps? If so, fill them using zoo::na.approx.

The byIQR method is an early implementation - it still has trouble picking up on all but the most extreme outliers.

Other preprocess: BindRawFlux, ConsoleDropOutliers, DropRainyDays, GenerateMetaTemplate, ImportMeta, ImportRawFlux, MetaDataImport, oldBindRawFlux, oldImportRawFlux

# As of 0.1.9 - just use byIQR = FALSE, it needs improvement
flux_data <- AutoDropOutliers(flux = flux.data, byIQR = FALSE)
# Check the log to see how many data points got dropped:
print(flux_data@log)