GEOXP

A Matlab Toolbox for Exploratory Spatial Analysis

To download.

Note: GeoXp exists also as an R package but is not yet downloadable because the help is incomplete)

Documentation (in french and for the R version)

updated, July 11th 2003

List of main functions:

angleplotmap.m : links a map and an angle plot (only the angle plot is active).

barmap.m : links a map and a bar plot.

boxplotmap.m : links a map and a box and whiskers plot.

clustermap.m : links a map and a bar map of a clustering variable (kmeans method).

dblebarmap.m : links a map and two bar plots.

dbledensitymap.m : links a map and two density estimators.

dblehistomap.m : links a map and two histograms.

densitymap.m : links a map and a density estimator.

driftmap.m : this function is meant for detecting trends

ginimap.m : links a map and a Gini plot (Lorentz curve).

histobarmap.m : links a map to an histogram and a bar plot.

histomap.m : links a map and an histogram.

mdsmap: links a map and a multidimensional scaling analysis

moranplotmap.m : links a map and a Moran scatterplot.

neighbourmap.m : links a map and a neighbour plot (scatterplot of variable against variable

for the neighboring sites)

pcamap.m: links a map and a scatterplot of principal axes of Principal Components Analysis

polyboxplotmap.m : links a map and a box and whiskers plots.

pprmap: links a map and two scatterplots of projections found by projection pursuit

scattermap.m : links a map and a two-dimensionnal scatterplot.

scatter3dmap: links a map and a three-dimensionnal scatterplot

sarmap.m : adjust a spatial autoregressive model on a subregion selected on the map.

semmap.m : adjust a spatial error model on a subregion selected on the map.

sirmap.m : links a map and one or two SIR scatterplots.

variocloudmap.m : links a map and a variogram cloud (only the variogram cloud is active).

List of demos:

loadsids: load the data for the demos

demoboxplot

demodensity

demogini

demohisto

demoscatter

demovario

List of auxiliary functions:

Some of the auxiliary functions come from the following other toolboxes:

- Classification: http://neural.cs.nthu.edu.tw/jang/matlab/toolbox/DCPR/

- Spatial Econometrics Toolbox : http://www.spatial-econometrics.com

- SmoothTlbxPC : http://www.unizh.ch/biostat/Software/smoothtoolbox

- Computational Statistics Toolbox : http://lib.stat.cmu.edu

- Strauss: http://www.biol.ttu.edu/Strauss/Matlab/matlab.htm

condmean : computes a conditional mean

contig.m : computes a contiguity matrix based on a threshold.

csppeda.m: projection pursuit (from Classification).

csppind: Chi square projection pursuit index (from Classification)

csppstrtrem: projection pursuit structure removal (from Computational Statistics Toolbox)

cvarpds.m : computes a weighted covariance matrix of a variable

eigen (from Strauss)

eucl (from Strauss)

find_neighbors.m : finds observations containing m nearest neighbors to each observation and returns an index to these neighboring observations. (from Spatial Econometrics Toolbox)

f_sar.m : evaluates concentrated log-likelihood for the spatial autoregressive model using sparse matrix algorithms. (from Spatial Econometrics Toolbox)

f_sem.m : evaluates concentrated log-likelihood for the spatial error model using sparse matrix algorithms. (from Spatial Econometrics Toolbox)

genpca.m: generalized principal components analysis

gini.m : computes a Gini index

hessian.m: computes finite difference hessian (from Spatial Econometrics Toolbox)

Ilocal.m : computes local Moran indices.

initkm2.m : Find the initial centers for a K-means clustering algorithm (from Classification).

invgen.m : computes the generalized inversion of a discrete cumulative distribution function.

invpd.m: dummy function to mimic Gauss invpd simply returns the inverse, with no checking for positivie definiteness (from Spatial Econometrics Toolbox)

kern_de1.m : Kernel estimator of the density (Parzen-Rosenblatt) for a unidimensional sample.

kern_re.m : Kernel estimator of regression (Nadaraya-Watson) for a bidimensional sample.

kmeans.m : Clustering based on kmeans method. (from Classification)

kmeans2.m : Clustering based on kmeans based on a distance matrix. (from Classification)

lndetfull.m: computes Pace and Barry's grid for log det(I-rho*W) using sparse matrices (from Spatial Econometrics Toolbox)

lndetmc.m : computes Barry and Pace MC approximation to log det(I-rho*W). (from Spatial Econometrics Toolbox)

lndetint.m: computes Pace and Barry's spline approximation to log det(I-rho*W) (from Spatial Econometrics Toolbox)

make_neighborsw.m : finds the nth nearest neighbor and constructs a spatial weight matrix based on this neighbor. (from Spatial Econometrics Toolbox)

metricclusters.m: computes a data-based metric adapted to the detection of clusters

metricoutliers.m: computes a data-based metric adapted to the detection of outliers

mds.m: non metric multidimensional scaling (from )

mdsfunc.m: objective function for mds (from )

moran.m : computes Moran's I-statistic for spatial correlation in the residuals of a regression model. (from Spatial Econometrics Toolbox)

nonormmoran.m: computes Moran's I-statistic for spatial correlation in the residuals of a regression model without normalizing the weight matrix

norm_cdf.m: computes the cumulative normal distribution for each component of x with mean m, variance v . (from Spatial Econometrics Toolbox)

norm_pdf : computes the normal probability density function for each component of x with mean m, variance v. . (from Spatial Econometrics Toolbox)

normw.m : normalize a spatial weight matrix to have row sums of unity. . (from Spatial Econometrics Toolbox)

noy.m : computes the triweigh kernel.

numbcla : computes edges of classes for a vector.

ols.m : least-square regression. (from Spatial Econometrics Toolbox)

quant.m : computes conditional quantiles

rotation.m : rotates objects of spatial coordinates type.

sar.m : computes spatial autoregressive model estimates y = p*W*y + X*b + e, using sparse matrix algorithms.

sem.m : computes spatial error model estimates y = XB + u, u = p*W*u + e, using sparse algorithms. (from Spatial Econometrics Toolbox)

selectmap.m : selects points on a map.

selectmapd.m : selects points on a map (demo version).

selectstat.m : selects objects on a statistic graph.

setdens.m : Callback function for the slider created in scattermap.m. Not usable outside this context.

setdens2.m : Callback function for the slider created in densitymap.m. Not usable outside this context.

setdens3.m : Callback function for the sliders created in dbledensitymap.m. Not usable outside this context.

setdens4.m : Callback function for the sliders created in sirmap.m. Not usable outside this context.

setdens5.m: Callback function for the slider created in angleplotmap.m

setw.m : Callback function for the slider created in morancontiplotmap.m. Not usable outside this context.

sirf.m : SIR method for a unidimentional output.

stepkm2.m : One step in k-means clustering (from Classification)

stdn_pdf : computes the standard normal probability density for each component of x with mean=0, variance=1. (from Spatial Econometrics Toolbox)

tostr.m: converts a vector of integer or real numbers to a text matrix (from Classification)

trilow: extracts the lower triangular portion (without diagonal) of a square symmetric matrix into a columnwise column vector. (from Strauss)

vecdist.m : Distance between two set of vectors.

vprgen.m : computes eigenvalues and eigenvectors in the SIR algorithm.