基于R语言地理探测器包（GD）空间异质性与驱动力分析

导入要分析的数据，并用切片器选择因变量和自变量，因变量在左侧，并删除无关数据，包括索引。

testdata <- read.csv(csvfile, header = TRUE, sep = ",")
testdata=testdata[3:9]

输入可供选择的分类方法

#3 approaches to calculate disgress.
  discmethod <- c("equal","natural","quantile")

输入可供选择的分类基本数量

#classify 4~6
discitv <- c(4:6)

筛选连续的自变量

#continuous variables
  continuous_variable <- colnames(testdata)[-1]

运行gdm函数 

#gdm function
  testgdm <- gdm(WSI ~ .,
                 continuous_variable = continuous_variable ,
                 data = testdata,
                 discmethod = discmethod, discitv = discitv)

分析结果

 完整代码

Sys.setlocale(category = 'LC_CTYPE', locale = 'C')#This is for using in pycharm.
library(GD)

filepath <- "E:/BaiduNetdiskDownload/Driving/Yl_naturalBreaks/"
setwd(filepath)#Change the workspace
temp=list.files(path=filepath,pattern="*.csv")#list csvfilenames in workspace
for( csvfile in temp){
  print(csvfile)
  testdata <- read.csv(csvfile, header = TRUE, sep = ",")

  #load data
  testdata <- read.csv("2003.csv", header = TRUE, sep = ",")
  testdata=testdata[3:9]

  #5 approach to calculate disgress.
  discmethod <- c("equal","natural","quantile")#select the best one among automatically 3 approaches of discrizetion 
  #classify 4~6 automatically
  discitv <- c(4:6)
  #continuous independent variables' name list
  continuous_variable <- colnames(testdata)[-1] #[-1] is for unselect dependent variable 

  #gdm function
  testgdm <- gdm(WSI ~ .,
                 continuous_variable = continuous_variable ,
                 data = testdata,
                 discmethod = discmethod, discitv = discitv)
  #dependent variable is WSI

  }

  print(testgdm)

}