RQADeforestation

agcube(filenames) Open the underlying tiff files via ArchGDAL. This opens all files and keeps them open. This has a higher upfront cost, but might lead to a speedup down the line when we access the data.

anti_diagonal_density(ts, thresh, metric)

Compute the average density of the diagonals perpendicular to the main diagonal for data series ts. Uses the threshold thresh and metric for the computation of the similarities.

countvalid(xout, xin)

Inner function to count the valid time steps in a datacube.
This function is aimed to be used inside of a mapCube call.

countvalid(cube)

Outer function to count the number of valid time steps in a cube.

countvalidag(xout, xin)

Inner function to count the valid time steps in a datacube.
This function is aimed to be used inside of a mapCube call.

gdalcube(indir, pol)

Load the datasets in indir with a polarisation pol as a ESDLArray. We assume, that indir is a folder with geotiffs in the same CRS which are mosaicked into timesteps and then stacked as a threedimensional array.

grouptimes(times, timediff=200000) Group a sorted vector of time stamps into subgroups where the difference between neighbouring elements are less than timediff milliseconds. This returns the indices of the subgroups as a vector of vectors.

Compute the forest masking thresholding and clustering of the rqadata in one step

rqatrend(xout, xin, thresh)

Compute the RQA trend metric for the non-missing time steps of xin, and save it to xout. thresh specifies the epsilon threshold of the Recurrence Plot computation

rqatrend(path::AbstractString; thresh=2, outpath=tempname()*".zarr")

Compute the RQA trend metric for the data that is available on path.

rqatrend(cube;thresh=2, path=tempname() * ".zarr")

Compute the RQA trend metric for the datacube cube with the epsilon threshold thresh.

rqatrend(xout, xin, thresh)

Compute the RQA trend metric for the non-missing time steps of xin, and save it to xout. thresh specifies the epsilon threshold of the Recurrence Plot computation

rqatrend_shuffle(cube; thresh=2, path=tempname() * ".zarr", numshuffle=300)

Compute the RQA trend metric for shuffled time series of the data cube cube with the epsilon threshold thresh for numshuffle tries and save it into path.

smooth(a, b, γ)

Weighted average of a and b with weight γ.

$(1 - γ) * a + γ * b$