距离权重改进的Pearson相关系数及应用

相关系数是表征两组数据相关性的度量指标。现有的相关系数仅计算数据本身在数值上的相关性，忽视了数据在地理分布上的特征，导致数据的规律性未被充分挖掘。在油气勘探领域，油气分布与空间信息密切相关。为此，采用Pearson相关系数的广义形式，结合空间权重的方法计算相关系数，并将不同搜索半径下的局部相关系数以折线图的形式研究两组数据的空间相关特征。模拟数据和实际数据的检测结果表明，空间权重改进的相关系数可以从空间上明确两组数据的潜在相关性。该方法有利于油气勘探的数据优选。

A Pearson correlation coefficient improved by spatial weight

The correlation coefficient is a kind of indicator for measuring the statistical relationship between two variables.Existing correlation coefficients calculate only the relationship in the values,but do not take in account of the variables' positions.So the rhythmicity of data is not fully explored.It is well known that the distribution of hydrocarbon is closely related to its geographical location in the hydrocarbon exploration.This paper combines generalized Pearson product-moment correlation coefficient with spatial weighs to calculate data local correlation,and analyze spatial correlation characteristics of two data sets with different searching radius.Based on synthetic and field data tests,the proposed approach demonstrates a spatial relationship between two variables,which is helpful for the hydrocarbon exploration.