Geographically Optimal Similarity (GOS) and the Third Law of Geography in R

 

Citation for package geosimilarity

To cite geosimilarity R package in publications, please use:

Song, Y. (2022) “Geographically Optimal Similarity”, Mathematical Geosciences. doi: 10.1007/s11004-022-10036-8.

 

 

1. Introduction to geosimilarity package

The package can be used to address following issues:

More details of GOS models can be found in Song (2022).

 

2. Spatial prediction using GOS model

According to Song (2022), GOS model consists of four primary steps: (1) Characterizing geographical configurations, (2) determining parameters for the optimal similarity, (3) spatial prediction using GOS and uncertainty assessment, and (4) model evaluation. The process of using geosimilarity package to conduct GOS modeling is presented as follows.

2.1 Characterizing geographical configurations

The geosimilarity package contains two spatial datasets:

install.packages("geosimilarity", dependencies = TRUE)
# or run the following code:
install.packages("geosimilarity", dep = TRUE)
library(geosimilarity)
## This is `geosimilarity` 3.8.
##                         
## To cite `geosimilarity` in publications, please use:
##                         
## Song, Y. (2022). Geographically Optimal Similarity. Mathematical Geosciences. doi: 10.1007/s11004-022-10036-8.
## 
data("zn")
head(zn)
## # A tibble: 6 × 12
##     Lon   Lat    Zn Elevation Slope Aspect  Water  NDVI   SOC    pH    Road  Mine
##   <dbl> <dbl> <dbl>     <dbl> <dbl>  <dbl>  <dbl> <dbl> <dbl> <dbl>   <dbl> <dbl>
## 1  120. -28.5    10      455. 0.236  306.  0.014  0.184 0.909  5.95 49.4     55.6
## 2  120. -28.4    30      451. 0.207  293.  2.20   0.202 0.906  6.05 49.0     51.1
## 3  120. -28.4    30      443. 0.285  325.  0.0119 0.163 0.848  5.76 45.1     45.0
## 4  120. -27.4    30      509. 0.236   98.4 3.06   0.204 0.851  5.82  0.0774  49.0
## 5  120. -28.3    33      427. 0.191  329.  3.53   0.179 0.933  5.85 39.9     39.8
## 6  120. -27.3    27      510. 0.211  105.  3.38   0.191 0.868  6.07  0.0468  48.7

Data pre-processing and variable selection:

# log-transformation
hist(zn$Zn)

zn$Zn <- log(zn$Zn)
hist(zn$Zn)


# remove outliers
k <- removeoutlier(zn$Zn, coef = 2.5)
## Remove 9 outlier(s)
dt <- zn[-k,]

# correlation
library("PerformanceAnalytics")
## Loading required package: xts
## Loading required package: zoo
## 
## Attaching package: 'zoo'
## The following objects are masked from 'package:base':
## 
##     as.Date, as.Date.numeric
## 
## Attaching package: 'PerformanceAnalytics'
## The following object is masked from 'package:graphics':
## 
##     legend
cor_dt <- dt[, c(3:12)]
chart.Correlation(cor_dt, histogram = TRUE, pch = 19)


# multicollinearity
library(car)
## Loading required package: carData
m1 <- lm(Zn ~ Slope + Water + NDVI + SOC + pH + Road + Mine, data = dt)
car::vif(m1)
##    Slope    Water     NDVI      SOC       pH     Road     Mine 
## 1.651039 1.232454 1.459539 1.355824 1.568347 2.273387 2.608347

In this step, the selected variables include Slope, Water, NDVI, SOC, pH, Road, and Mine.

2.2 Determining the optimal similarity

In the gos_bestkappa() function, if you set more optional numbers to the kappa vector and a higher value of the cross-validation repeat times nrepeat, a \(\kappa\) value enabling more accurate prediction will be selected, but the computation time will be increased. You can specify the cores parameter to use multiple CPU cores for parallel computing.

The default ratio of train set to test set in gos_bestkappa() is 1:1(0.5). You can specify the ratio of train set to test set by nsplit parameter

system.time({
b1 <- gos_bestkappa(Zn ~ Slope + Water + NDVI  + SOC + pH + Road + Mine,
                    data = dt,
                    kappa = c(0.01, 0.05, 0.1, 0.2, 0.5, 1),
                    nrepeat = 2,
                    cores = 1)
})
##    user  system elapsed 
##    5.24    0.07    5.46
b1$bestkappa
## [1] 0.1
b1$cvmean
## # A tibble: 6 × 2
##   kappa  rmse
##   <dbl> <dbl>
## 1  0.01 0.681
## 2  0.05 0.663
## 3  0.1  0.659
## 4  0.2  0.661
## 5  0.5  0.664
## 6  1    0.665

system.time({
b2 <- gos_bestkappa(Zn ~ Slope + Water + NDVI  + SOC + pH + Road + Mine,
                    data = dt,
                    kappa = c(seq(0.01, 0.1, 0.01), seq(0.2, 1, 0.1)),
                    nrepeat = 10,
                    cores = 6)
})
##    user  system elapsed 
##    0.05    0.07   21.69
b2$bestkappa
## [1] 0.08
b2$cvmean
## # A tibble: 19 × 2
##    kappa  rmse
##    <dbl> <dbl>
##  1  0.01 0.694
##  2  0.02 0.680
##  3  0.03 0.674
##  4  0.04 0.672
##  5  0.05 0.669
##  6  0.06 0.667
##  7  0.07 0.667
##  8  0.08 0.667
##  9  0.09 0.667
## 10  0.1  0.667
## 11  0.2  0.669
## 12  0.3  0.671
## 13  0.4  0.672
## 14  0.5  0.672
## 15  0.6  0.673
## 16  0.7  0.673
## 17  0.8  0.673
## 18  0.9  0.673
## 19  1    0.673

library(cowplot)

plot_grid(b1$plot,b2$plot,nrow = 1,label_fontfamily = 'serif',
          labels = paste0('(',letters[1:2],')'),
          label_fontface = 'plain',label_size = 10,
          hjust = -1.5,align = 'hv')
Figure 1. Processes of determining the optimal similarity. (a) The optional kappa is (0.01, 0.05, 0.1, 0.2, 0.5, 1) and nrepeat is 2. (b) The optional kappa is (0.01, 0.02, …, 0.09, 0.1, 0.2, …, 1) and nrepeat is 10.
Figure 1. Processes of determining the optimal similarity. (a) The optional kappa is (0.01, 0.05, 0.1, 0.2, 0.5, 1) and nrepeat is 2. (b) The optional kappa is (0.01, 0.02, …, 0.09, 0.1, 0.2, …, 1) and nrepeat is 10.

2.3 Spatial prediction

system.time({
g2 <- gos(Zn ~ Slope + Water + NDVI  + SOC + pH + Road + Mine,
          data = dt, newdata = grid, kappa = 0.08, cores = 6)
})
##    user  system elapsed 
##    0.09    0.04    7.13
grid$pred <- exp(g2$pred)
grid$uc99 <- g2$`uncertainty99`

library(ggplot2)
library(viridis)
## Loading required package: viridisLite

f1 = ggplot(grid, aes(x = Lon, y = Lat, fill = pred)) +
  geom_tile() +
  scale_fill_viridis(option="magma", direction = -1) +
  coord_equal() +
  labs(fill='Prediction') +
  theme_bw()
f2 = ggplot(grid, aes(x = Lon, y = Lat, fill = uc99)) +
  geom_tile() +
  scale_fill_viridis(option="mako", direction = -1) +
  coord_equal() +
  labs(fill=bquote(Uncertainty~(zeta==0.99))) +
  theme_bw()

plot_grid(f1,f2,nrow = 1,label_fontfamily = 'serif',
          labels = paste0('(',letters[1:2],')'),
          label_fontface = 'plain',label_size = 10,
          hjust = -1.5,align = 'hv')
Figure 2. Geographially optimal similarity (GOS)-based prediction (a) and uncertainty (b).
Figure 2. Geographially optimal similarity (GOS)-based prediction (a) and uncertainty (b).

In addition, the following codes can be used to plot uncertainty under different \(\zeta\) values.

uc <- g2 %>%
  dplyr::select(dplyr::starts_with("uncertainty")) %>%
  dplyr::bind_cols(grid[,2:3],.) %>%
  tidyr::pivot_longer(cols = -c(1,2),
                      names_to = "uncertainty",
                      values_to = "value")
ggplot(uc, aes(x = Lon, y = Lat, fill = value)) +
  geom_tile() +
  scale_fill_viridis(option="mako", direction = -1) +
  coord_equal() +
  facet_wrap(~ uncertainty) +
  labs(fill='Uncertainty') +
  theme_bw()
Figure 3. Geographially optimal similarity (GOS)-based spatial prediction uncertainties under different \zeta values.
Figure 3. Geographially optimal similarity (GOS)-based spatial prediction uncertainties under different \(\zeta\) values.

2.4 Model evaluation

We can compare model accuracy of GOS with various models, such as kriging, multivariate regression, regression kriging, random forest, BCS, etc., as shown in Song (2022). Here is a simple example of comparing modeling accuracy between BCS and GOS.

set.seed(99)
# split data for validation: 50% training; 50% testing
split <- sample(1:nrow(dt), round(nrow(dt)*0.5))
train <- dt[split,]
test <- dt[-split,]

library(DescTools)
## 
## Attaching package: 'DescTools'
## The following object is masked from 'package:car':
## 
##     Recode
# BCS
h1 <- gos(Zn ~ Slope + Water + NDVI  + SOC + pH + Road + Mine,
          data = train, newdata = test, kappa = 1)
MAE(test$Zn, h1$pred)
## [1] 0.5158373
RMSE(test$Zn, h1$pred)
## [1] 0.6599409

# GOS
h2 <- gos(Zn ~ Slope + Water + NDVI  + SOC + pH + Road + Mine,
          data = train, newdata = test, kappa = 0.08)
MAE(test$Zn, h2$pred)
## [1] 0.5089462
RMSE(test$Zn, h2$pred)
## [1] 0.6523436

As a result, the MAE of BCS is 0.5158 and the MAE of GOS is 0.5089, the RMSE of BCS is 0.6599 and the RMSE of GOS is 0.6523. Compared with BCS, GOS reduced 1.34% of MAE and 1.15% of RMSE.