Add trend line to a scatter plot

Question

Beatriz Sanchez 2018-9-27

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/421072-add-trend-line-to-a-scatter-plot

编辑： Adam Danz 2020-4-5

Hi, I'm ploting some data in a scatter and I want to add a trending line to the plot, but the imput x and y for the scatter are vectors and the fit function take x and y as matrix, how can I do this, my code below:

vec= [-180, -90, 0, 90, 180]
x_mean= randi (1000, 3, 5)
for i= 1:3
hold on
scatter (vec, x_mean (i, :), 'b')
end

I want to add one trend line that consider ALL of the values in the scatter. Need help. Thanks

2 个评论
显示无隐藏无

Adam Danz 2018-9-27

A trend line (like lsline) is usually understood as a regression line and it would definitely not connect all of the values in the scatter. So, is a trend line really what you want?

You could use plot(vect, x_mean) to connect the series of lines but that would result in 3 lines since there are 3 rows in x_mean.

Beatriz Sanchez 2018-9-27

I don't want to exactly connect them, I want a trend line that consider all of them

请先登录，再进行评论。

请先登录，再回答此问题。

Answer 1

Adam Danz 2018-9-27

1
链接

此回答的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/421072-add-trend-line-to-a-scatter-plot#answer_338738

编辑：Adam Danz 2020-4-5

在 MATLAB Online 中打开

Use coef = polyfit(x,y) to compute the regression coefficients. Then use refline(coef(1),coef(2)) to plot the regression line.

 coef = polyfit(repmat(vec, size(x_mean,1), 1), x_mean, 1);
 h = refline(coef(1), coef(2));

Alternatively, you can use lsline(ax) to add the least squares regression line to each set of data within the axes without needing to compute the regression coefficients.