nei.efdr

Given an image, this function first computes the 2d DWT and then returns a 
matrix of size <code>N</code> by <code>b</code> where <code>N</code> is the number of wavelets and <code>b</code> 
is the number of neighbours per wavelet. Two wavelets are deemed
to be neighbours according to the metric of Shen, Huang and Cressie (2002). The distance metric is a function of the
spatial separation, the resolution and the orientation.

Enhanced False Discovery Rate (EFDR) is a tool to detect anomalies
in an image. The image is first transformed into the wavelet domain in
order to decorrelate any noise components, following which the coefficients
at each resolution are standardised. Statistical tests (in a multiple
hypothesis testing setting) are then carried out to find the anomalies. The
power of EFDR exceeds that of standard FDR, which would carry out tests on
every wavelet coefficient: EFDR choose which wavelets to test based on a
criterion described in Shen et al. (2002). The package also provides
elementary tools to interpolate spatially irregular data onto a grid of the
required size. The work is based on Shen, X., Huang, H.-C., and Cressie, N.
'Nonparametric hypothesis testing for a spatial signal.' Journal of the
American Statistical Association 97.460 (2002): 1122-1140.

Andrew ZammitMangion

EFDR

Wavelet-Based Enhanced FDR for Detecting Signals from Complete
or Incomplete Spatially Aggregated Data

Andrew Zammit-Mangion

Hsin-Cheng Huang

nei.efdr function

<dl><dt>Z</dt>
<dd>image of size <code>n1</code> by <code>n2</code> where both <code>n1,n2</code> have to be powers of two</dd>
<dt>wf</dt>
<dd>type of wavelet to employ. Please see <code>waveslim::wave.filter</code> for a full list of wavelet names</dd>
<dt>J</dt>
<dd>number of resolutions to employ in the wavelet decomposition</dd>
<dt>b</dt>
<dd>number of neighbours to consider in EFDR</dd>
<dt>parallel</dt>
<dd>number of cores to use with parallel backend; needs to be an integer less than the number of available cores</dd></dl>

Arguments

Find wavelet neighbourhood — nei.efdr

<dl>

<dt>Z</dt>
<dd>image of size <code>n1</code> by <code>n2</code> where both <code>n1,n2</code> have to be powers of two</dd>


<dt>wf</dt>
<dd>type of wavelet to employ. Please see <code>waveslim::wave.filter</code> for a full list of wavelet names</dd>


<dt>J</dt>
<dd>number of resolutions to employ in the wavelet decomposition</dd>


<dt>b</dt>
<dd>number of neighbours to consider in EFDR</dd>


<dt>parallel</dt>
<dd>number of cores to use with parallel backend; needs to be an integer less than the number of available cores</dd>

</dl>

nei.efdr: Find wavelet neighbourhood

Description

Usage

Value

Arguments

References

Examples