xdem.spatialstats#
Spatial statistical tools to estimate uncertainties related to DEMs
Functions
|
Convolution on a number n_N of 2D images of size N1 x N2 using a number of kernels n_M of sizes M1 x M2, using either scipy.signal.fftconvolve or accelerated numba loops. |
|
Construct the spatial correlation function from a dataframe of variogram parameters. |
|
Construct the spatial covariance function from a dataframe of variogram parameters. |
|
Estimate and model the heteroscedasticity (i.e., variability in error) according to a list of explanatory variables from a proxy of differenced values (e.g., elevation differences), if possible compared to a source of higher precision. |
|
Estimate and model the spatial correlation of the input variable by empirical variogram sampling and fitting of a sum of variogram model. |
|
Fit a sum of variogram models to an empirical variogram, with weighted least-squares based on sampling errors. |
|
Get per-bin array statistic for a list of array input variables, based on the results of an independent N-D binning. |
|
Construct the sum of spatial variogram function from a dataframe of variogram parameters. |
|
Infer heteroscedasticity from differenced values on stable terrain and a list of explanatory variables. |
|
Infer spatial correlation of errors from differenced values on stable terrain and a list of variogram model to fit as a sum. |
|
Estimate an interpolant function for an N-dimensional binning. |
|
Apply a mean filter to an image with a square or circular kernel of size p and with NaN values ignored. |
|
N-dimensional binning of values according to one or several explanatory variables with computed statistics in each bin. |
|
Number of effective samples derived from numerical integration for any sum of variogram models over a circular area. |
|
Number of effective samples approximated from exact disk integration of a sum of any number of variogram models of spherical, gaussian, exponential or cubic form over a disk of a certain area. |
|
Exact number of effective samples derived from a double sum of covariance with euclidean coordinates based on |
|
Approximated number of effective samples derived from a double sum of covariance subsetted on one of the two sums, based on euclidean coordinates with the provided variogram parameters. |
|
Calculate the normalized median absolute deviation (NMAD) of an array. |
|
Compute the number of effective samples, i.e. the number of uncorrelated samples, in an area accounting for spatial correlations described by a sum of variogram models. |
|
Monte Carlo patches method that samples multiple patches of terrain, square or circular, of a certain area and computes a statistic in each patch. |
|
Plot a statistic and its count along a single binning variable. |
|
Plot one statistic and its count along two binning variables. |
|
Plot empirical variogram, and optionally also plot one or several model fits. |
|
Sample empirical variograms with binning adaptable to multiple ranges and spatial subsampling adapted for raster data. |
|
Spatial propagation of elevation errors to an area using the estimated heteroscedasticity and spatial correlations. |
|
Standardize the proxy differenced values using the modelled heteroscedasticity, re-scaled to the spread statistic, and generate the final standardization function. |
Classes
|