setProj: Set Up an ESTREND Project
In USGS-R/restrend: EStimate TRENDs

setProj

R Documentation

Set Up an ESTREND Project

Description

Define the data and other characteristics of an ESTREND project.

Usage

setProj(project, data, STAID, DATES, Snames, FLOW = NULL, Covars = NULL,
  type = "seasonal", Start = NULL, End = NULL, tol = NULL,
  min.obs = 20)

Arguments

`project`	the name of the project to set up. The project name is forced to all lower case.
`data`	the dataset to use in for the project.
`STAID`	the name of the column in `data` that contains the station identifiers. The data are forced to character for the analysis.
`DATES`	the name of the column in `data` containing the sample dates. This column must be class "Date" for seasonal analyses, but can be numeric or integer for annual analyses.
`Snames`	the name of the columns in `data` containing the sample data for trend analysis. These must be of class "numeric," "integer," "lcens," or "qw."
`FLOW`	the name of the column in `data` containing the streamflow for each sample.
`Covars`	the name of the columns in `data` containing any covariate data for trend analysis.
`type`	the kind of analysis. Must be "seasonal," "tobit," "annual," or "monthly." Only the fist letter is necessary. See Details.
`Start`	the starting date for the analysis. For seasonal analyses, must be "Date" or a character string the represents a date. For annual analyses, must match the type of `DATES`. See Details.
`End`	the ending date for the analysis. For seasonal analyses, must be "Date" or a character string the represents a date. For annual analyses, must match the type of `DATES`. To guarantee that the periods are set up correctly, `End` should be the first day of the month following the actual last day.
`tol`	the tolerance for the samples dates. To be included in a regular analysis, the first sample within the `Start` to `End` time period must be within `tol` and likewise for the last sample. if `tol` is `NULL`, then it is set to 5 percent of the time frame, except for annual series analysis when it is set to 1 year.
`min.obs`	the minimum number of observations required for a trend analysis.

Details

If Start and End are NULL, then a ragged analysis is set up, and each variable set up for each station is analyzed on the available data. Otherwise, the analysis is regular and will be restricted to the time frame specified by Start and End.

If type is "seasonal," then the data are processed for a seasonal Kendall type of analysis—seasons are defined (12, 6, 4, and 3 per year) and evaluated to select the "best" number of seasons. If, type is "monthly," then the data are processed for a seasonal Kendall type of analysis—seasons defined as months, but the number of seasons is set by the months during which sampling occured. There must be at least 1/2 of possible samples in a month to be included. If type is "tobit" or "annual," then no seasonal analysis is done because the data are ready for analysis. If type is "annual," then the data must be uncensored.