SAAR
Spatial Analysis using ArcGIS Engine and R
1. Instruction
Example Video
Introducing SAAR
2. Tutorials
Tutorial 1. Explore Data Analysis (EDA)
SAAR contains EDA tools with the support of dynamic linking and brushing techniques among statistical graphs and a layer map in a map view. EDA tools support not only generic graphics such as histogram, boxplot, and scatterplot but also violin and quantile-comparison plot.
Histogram, Boxplot, Scatter plot(upper) and Violin, Quantile-comparison plot (bottom).
Tutorial 2. Spatial autocorrelation and Exploratory spatial data analysis (ESDA)
SAAR provides a set of global and local spatial autocorrelation measures, and ESDA tools, including Conditioned Choropleth maps, a connectivity histogram and map, a Moran scatter plot, and a spatial correlogram.
Distribution of House Values
1) Global Spatial Autocorrelation
2) Local Spatial Autocorrelation
3) Moran scatter plot and spatial correlogram
Moran scatter plot (left) and spatial correlogram(right)
Tutorial 3. Data transformations and regression analysis
SAAR furnishes a Box-Cox transformation tool to convert an input variable to one that better mimics a normal distribution (i.e., a bell-shaped curve). Furthermore, SAAR supports various regression analysis tools, including simple linear regression, generalized linear model (GLM), various spatial regressions, and Moran eigenvector spatial filtering (MESF) for linear and GLM specifications. In addition,
1) Box-Cox transformation
Normal Q-Q plot (left) and Transformed Q-Q plot (right)
2) Linear regression
3) Spatial regression
4) Moran eigenvector spatial filtering (ESF)
5) Comparison of three models using residual distribution
Linear regression, Spatial regression, Moran eigenvector spatial filtering (MESF)
Tutorial 4. Geovisualization for spatial data uncertainty
SAAR provides toolsets to explore uncertainty in spatial data and SDA output in two different contexts: geovisualization, and map classification. In this lab, we will use uncertainty geovisualization tools developed based on bivariate mapping technique.
1) Uncertainty in spatial data
3. Critics
1) Advantage
2) Disadvantage
