readme.txt

数据挖掘算法中空间同位规则挖掘的matlab源码

data:
sample.data : a sample data which consists of 74 instances and 10 features. Format: (feature, x, y)
featurei.txt: instances of feature i (i=0 to 9)


Scripts and functions:
vis : plot all the instances in a map

join: 
      self join of sample data to produce a table with the format: (id1,feature1,x1,y1,id2,feature2,x2,y2,distance)
      the table (matrix) name is sample_size_2

part_index(sample_size_2,sample,thrd):
      this function calculates the pairwise co-location with participation index and participation ratios (pr)
      input:
          sample_size_2 is the table produced by join
          sample is the original sample data
          thrd is the distance threshold to define neighborhood
      output: 
          size 2 colocation with the format: (participation index, feature1, feature2, pr for feature1, pr for feature2)

sortrows(part)
      sort a matrix, can be used to sort the size 2 co-location table

prev(part, prev_thrd);
      this function selects a subset of size 2 co-location with participation index greater than prev_thrd 
      input:
          part is the size 2 co-location table
          prev_thrd is the participation index threshold
      out: 
          the prevalent size 2 co-locations
         
plot_prev(sample, prev_part):
      this function plots the prevalent size 2 co-location in same color and all instances from other feature in blue
      input: 
          sample is the original sample data
          prev_part is the output of the prev(part,prev_thrd) function