Spatial cold hot spot analysis

Spatial cold hot spot analysis

Definition and principle

Hotspot analysis tools can calculate Getis-Ord Gi * statistics (called Gi-asterisks) for each element in the data set. With the obtained z-score and p-value , you can know the spatial clustering of high-value or low-value features. The way this tool works is: View every feature in the environment of nearby features. High-value elements tend to attract attention, but may not be a hot spot with statistical significance. To become a hot spot with significant statistical significance, the elements should have high values ​​and be surrounded by other elements that also have high values. The partial sum of an element and its neighbors will be compared with the sum of all elements; when the partial sum is so different from the expected partial sum that it cannot be a randomly generated result, a statistically significant Learning the z-score . If FDR correction is applied , the statistical significance is adjusted according to multiple tests and spatial dependence.

Hot spot analysis is based on the idea of ​​zero hypothesis testing commonly used in statistical inference. Because our eyes and brain are all patterns behind analyzing data all the time, even randomly distributed events may show a certain degree of agglomeration in space. The goal of hotspot analysis tools is to identify areas with statistically significant clustering. Because this shows that these events are being affected by certain spatial process factors, there is a spatial correlation.

Definition from argis official website

Let's look at the actual operation

The data preparation and the normal spatial distribution are consistent, such as

Insert picture description here

Ordinary spatial distribution map-you can see where is high and where is low. But if you want quantitative analysis, then you need to use hot spot analysis tools

First, let's see if he is random globally, accumulated or discrete

因此First carry out spatial autocorrelation analysis

Insert picture description here
The options are as follows

Insert picture description here

Results: with spatial clustering characteristics

Therefore, space cold spot analysis can be performed

Insert picture description here

Options

Insert picture description here

The result is the following figure

After typography and font adjustment

Insert picture description here
Code words are not easy, if they are useful, please like or follow! !

Published 11 original articles · won 23 · visited 9469

Guess you like

Origin blog.csdn.net/weixin_43897615/article/details/105540502