95% confidence intervalsof regression line!

2017 11 1
1 个回答
回答已采纳
更新时间：2017 12 20
6 次查看（30 天）

0 个投票

Hi! How can I calculate the 95% confidence intervals(numerically) of a regression line after using the the command regress?

请先登录，再回答此问题。

 采纳的回答

Star Strider
Star Strider 2017-11-1

0 个投票

I like regress, even with its limitations. However, it would probably be best for you to re-run your regression with the fitlm (link) function, then use the predict (link) function to calculate the confidence intervals. The ‘Xnew’ in the documentation are your existing independent variable values.

4 个评论
显示 2更早的评论 隐藏 2更早的评论

Panos Ale
Panos Ale 2017-12-19
I know it's has been a while ,I used the fitlim and predict and everything is ok!I know it's has been a while but can you also clarify to how can I calculate the 95% CI of the slope as two numbers(upper and lower) because I want to plot them two error bars. I hope I make myself understandable. thanks for the help anyway :)
Star Strider
Star Strider 2017-12-19
The regress function will calculate the parameter confidence limits. You can then plot them any way you like.
As I mentioned in my original Answer, use fitlm and predict to calculate the confidence intervals on the regression line.
Panos Ale
Panos Ale 2017-12-20
编辑：Panos Ale 2017-12-20
Clear and helpful as always!Thank you very much!
Star Strider
Star Strider 2017-12-20
As always, my pleasure!

请先登录，再进行评论。

更多回答（0 个）

类别

在 帮助中心 和 File Exchange 中查找有关 Linear Regression 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by