Comparación Indice de Moran R y Geoda

Se hara una comparación entre dos softwares para la autocorrelación espacial calculando el Indice de Moran.

library(readxl)
df <- read_excel("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/BD_MODELADO.xlsx")
library(DT)
datatable(df, class='cell-border stripe', filter='top', options = list(pageLength=10,autoWidth=T))

library(ggplot2)

## Warning: package 'ggplot2' was built under R version 4.0.3

ggplot(df, aes(x = Avg_X_MCB, y=Avg_Y_MCE, colour=Avg_CEa_07))+
   geom_point(size = 5)+
   scale_color_continuous(type = 'viridis')

ggplot(df, aes(x = Avg_X_MCB, y=Avg_Y_MCE, colour=Avg_CEa_15))+
   geom_point(size = 5)+
   scale_color_continuous(type = 'viridis')

ggplot(df, aes(x = Avg_X_MCB, y=Avg_Y_MCE, colour=NDVI))+
   geom_point(size = 5, shape = 15)+
   scale_color_continuous(type = 'viridis')

INDICE DE MORAN

library(ape)
t.dists<-as.matrix(dist(cbind(df$Avg_X_MCB, df$Avg_Y_MCE)))
dim(t.dists)

## [1] 313 313

t.dists.inv<-1/t.dists
t.dists.inv[is.infinite(t.dists.inv)] <- 0
diag(t.dists.inv)<-0
t.dists.inv[1:5, 1:5]

##            1          2          3          4          5
## 1 0.00000000 0.19320482 0.02207833 0.05403989 0.04558763
## 2 0.19320482 0.00000000 0.02476496 0.04650837 0.05738726
## 3 0.02207833 0.02476496 0.00000000 0.01665161 0.03039597
## 4 0.05403989 0.04650837 0.01665161 0.00000000 0.03392139
## 5 0.04558763 0.05738726 0.03039597 0.03392139 0.00000000

Indice de Moran de conductividad electrica a 70 cm de profundidad

I.Moran_70<-Moran.I(df$Avg_CEa_07, t.dists.inv); I.Moran_70

## $observed
## [1] 0.2687468
## 
## $expected
## [1] -0.003205128
## 
## $sd
## [1] 0.004665906
## 
## $p.value
## [1] 0

Indice de Moran de conductividad electrica a 150 cm de profundidad

I.Moran_150<-Moran.I(df$Avg_CEa_15, t.dists.inv); I.Moran_150

## $observed
## [1] 0.160951
## 
## $expected
## [1] -0.003205128
## 
## $sd
## [1] 0.00465455
## 
## $p.value
## [1] 0

Indice de Moran del NDVI

I.Moran_NDVI<-Moran.I(df$NDVI, t.dists.inv); I.Moran_NDVI

## $observed
## [1] 0.09750403
## 
## $expected
## [1] -0.003205128
## 
## $sd
## [1] 0.004644979
## 
## $p.value
## [1] 0

GeoDa

Indice de Moran de conductividad electrica a 70 cm de profundidad

knitr::include_graphics("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/GEODA/BD_MODELADO - GEODA7.png")  

Permutacional

knitr::include_graphics("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/GEODA/hist.png") 

knitr::include_graphics("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/GEODA//75_dist_wy.png") 

knitr::include_graphics("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/GEODA/lisa_75.png") 

Indice de Moran de conductividad electrica a 150 cm de profundidad

knitr::include_graphics("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/GEODA/BD_MODELADO - GEODA150.png")  

knitr::include_graphics("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/GEODA/lisa 150.png") 

knitr::include_graphics("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/GEODA/150.png") 

Indice de Moran NDVI

knitr::include_graphics("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/GEODA/BD_MODELADO - GEODA_NDVI.png") 

knitr::include_graphics("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/GEODA/ndvi clus.png") 

knitr::include_graphics("C:/Users/Sofia Hernandez/Documents/2020-2/Computacion estadistica/GEODA/ndvi lisa.png") 

Se evidencia una autocorrelación en ambos metodos usados