calc_dfpart

calculate degrees of freedom for partitioned GLS

These tools were created to test map-scale hypotheses about trends in large
remotely sensed data sets but any data with spatial and temporal variation
can be analyzed. Tests are conducted using the PARTS method for analyzing spatially
autocorrelated time series
(Ives et al., 2021: <doi:10.1016/j.rse.2021.112678>).
The method's unique approach can handle extremely large data sets that other
spatiotemporal models cannot, while still appropriately accounting for
spatial and temporal autocorrelation. This is done by partitioning the data
into smaller chunks, analyzing chunks separately and then combining the
separate analyses into a single, correlated test of the map-scale hypotheses.

Clay Morrow

remotePARTS

Spatiotemporal Autoregression Analyses for Large Data Sets

Anthony Ives

calc_dfpart function

<dl><dt>partsize</dt>
<dd>number of pixels in each partition</dd>
<dt>p</dt>
<dd>number of predictors in alternate model</dd>
<dt>p0</dt>
<dd>number of parameters in null model</dd></dl>

Arguments

calculate degrees of freedom for partitioned GLS — calc_dfpart

<dl>

<dt>partsize</dt>
<dd>number of pixels in each partition</dd>


<dt>p</dt>
<dd>number of predictors in alternate model</dd>


<dt>p0</dt>
<dd>number of parameters in null model</dd>

</dl>

calc_dfpart: calculate degrees of freedom for partitioned GLS

Description

Usage

Value

Arguments