Reading Robustfit Stats Results

8 次查看（过去 30 天）

balsip 2017-5-13

2
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/340159-reading-robustfit-stats-results

评论： John McDowell 2018-12-3

Using robustfit, I'm looking to find the R^2 and p-value for the resulting linear regression. I'm using two vectors, and returning stats. The code takes the form:

[X,STATS] = robustfit(A,B)

Is anything in the resulting STATS structure the R^2 or the p-value?

Clearly, the p listed within STATS is a tantalizing possibility, but why is it a 2x1 matrix? I would expect it to be a single value; one R^2 value for the linear regression.

The Matlab documentation for robustfit ( https://www.mathworks.com/help/stats/robustfit.html ) tells me that p is the "p-values for t", and t is the "Ratio of b to se", and you can see how a newbie to stats might be perplexed at this point.

I do not have any additional toolboxes, FYI.

Full disclosure: I am slowly getting familiar with statistics in general, so please excuse the shortcomings and oversights.

Thanks, everyone.

1 个评论
显示 -1更早的评论隐藏 -1更早的评论

John McDowell 2018-12-3

Did you ever work out the answer to this question? I am trying to work out the similarity between two curves, and had been recommended robusfit(). However, the 2x1 matrix of pvalues confuses me also, and I can find no plainly worded explanation of the STAT outputs.

请先登录，再进行评论。

请先登录，再回答此问题。