How to fit line of best fit to scatterplot in R2012a

Question

Bianca Elena Ivanof 2016-11-18

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/312916-how-to-fit-line-of-best-fit-to-scatterplot-in-r2012a

评论： Star Strider 2016-11-19

Hello,

Suppose I have 2 variables

curiosity = [4.916666667
3.916666667
3.666666667
5.083333333
4.666666667
3.75];
prediction = [19.58680175
-8.006943896
-2.934021031
32.79861546
78.37847569
41.84028306]; 
scatter(curiosity,prediction)

and I want to fit a standard regression line to it (without changing between polynomials - I'm only interested in the linear trend and not the cubic, quadratic or quartic one), how should I proceed? I've tried out solutions provided in a couple of answers to previous MATLAB questions but been unsuccessful so far.

Thank you very much in advance.

0 个评论
显示 -2更早的评论隐藏 -2更早的评论

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Star Strider 2016-11-18

0
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/312916-how-to-fit-line-of-best-fit-to-scatterplot-in-r2012a#answer_243835

在 MATLAB Online 中打开

Two possible ways:

Regression_Coefficients = polyfit(curiosity,prediction,1);

or:

Regression_Coefficients = [curiosity, ones(size(curiosity))]\prediction;
Regression_Coefficients =
    21.8212e+000   -67.6145e+000
Regression_Coefficients =
    21.8212e+000
   -67.6145e+000

2 个评论
显示无隐藏无

Bianca Elena Ivanof 2016-11-19

hello and thank you very much,

this gives me the regression coefficients but it doesn't fit a line to my scatterplot - or maybe I don't know how to do this still?

Star Strider 2016-11-19

在 MATLAB Online 中打开

My pleasure.

To plot the line for each (they both give the same result), the full code becomes:

curiosity = [4.916666667
3.916666667
3.666666667
5.083333333
4.666666667
3.75];
prediction = [19.58680175
-8.006943896
-2.934021031
32.79861546
78.37847569
41.84028306];
Regression_Coefficients_P = polyfit(curiosity,prediction,1)
Regression_Coefficients_M = [curiosity, ones(size(curiosity))]\prediction
x_plot = [min(curiosity); max(curiosity)];                      % Only Need Two Points To Plot The Linear Fit
y_plot_P = polyval(Regression_Coefficients_P, x_plot)           % Use ‘polyval’
y_plot_M = [x_plot [1; 1]] * Regression_Coefficients_M          % Use Matrix Algebra
figure(1)
scatter(curiosity, prediction)
hold on
plot(x_plot, y_plot_P)
hold off
grid
figure(2)
scatter(curiosity, prediction)
hold on
plot(x_plot, y_plot_M)
hold off
grid