Skip to content

Latest commit

 

History

History

Rank-sum_ratio(RSR)

Folders and files

NameName
Last commit message
Last commit date

parent directory

..
 
 
 
 

秩和比综合分析法Rank-sum ratio(RSR)

What:

Rank-sum ratio (RSR method), proposed by Chinese scholar Tian Fengtiao of the Chinese Academy of Preventive Medicine in 1988, is a statistical analysis that combines the advantages of classical parameter statistics and modern non-parametric statistics. It is not only suitable for the comprehensive evaluation of four-grid table data, but also for the comprehensive evaluation of row×list data, and also for the comprehensive evaluation of measurement data and classified data.
The basic idea is to obtain a dimensionless statistic RSR in a matrix of n rows (n evaluation objects) and p columns (p evaluation indicators) through rank conversion, and rank or classify the evaluation objects according to the RSR value. Sort. The value of the rank-sum ratio can contain the information of all evaluation indicators and show the comprehensive level of these evaluation indicators. The larger the RSR value, the better the comprehensive evaluation.

Why:

Advantages of RSR:

  1. The rank sum ratio is a new statistic, a carrier of compound information, with large capacity and strong plasticity.
  2. The rank-sum ratio method is a new broad-spectrum practical quantitative method, or statistical information method, which integrates parametric statistics and non-parametric statistics. It has description and inference. It can improve the level of statistical analysis and re-analysis and satisfy people. There are various demands in statistical research and statistical management. Therefore, we say that the rank-sum ratio method is an innovation of quantitative methods and has extremely broad development prospects.
  3. The key step of the rank-sum ratio method is rank substitution, which has powerful statistical information functions, strong pertinence, flexibility, simple operation and high application value.
  4. Through transplantation and grafting. Using the obtained rank sum ratio, you can reach in all directions. This method occupies an important position in quantitative research. There are details in roughness and roughness in fineness. The rules and characteristics are natural and clear. It is an effective means for comparison and finding relationships.

Disadvantages of RSR:

  1. The disadvantage of the rank-sum ratio evaluation method is that the main basis for ranking is the rank of the original data. The final calculated RSR value reflects the difference in the overall rank, and has nothing to do with the degree of the difference between the original data. When the indicator is converted to rank, some information of the original data will be lost, such as the size difference of the original data.

How:

The code had been written in Rank-sum ratio(RSR).py .

References:

https://baike.baidu.com/item/%E7%A7%A9%E5%92%8C%E6%AF%94%E6%B3%95/2805901?fr=aladdin
https://zhuanlan.zhihu.com/p/38209882
https://baike.baidu.com/item/%E7%A7%A9%E5%92%8C%E6%AF%94%E6%B3%95/2805901#1