how to create two bell shape curves

2 ビュー (過去 30 日間)
jenka
jenka 2013 年 4 月 24 日
with the areas under these curves equal to 1, the same mean but different standard deviations. Thanks!

採用された回答

Wayne King
Wayne King 2013 年 4 月 24 日
編集済み: Wayne King 2013 年 4 月 24 日
Do you have the Statistics Toolbox?
x = -10:0.01:10;
y = normpdf(x,0,1);
y1 = normpdf(x,0,sqrt(2));
If you do not have the Statistics Toolbox, you can just use the definition of the Gaussian
f = @(x) 1/sqrt(2*pi)*exp(-x.^2/2);
integral(f,-10,10)
g = @(x) 1/sqrt(2*pi*2)*exp(-x.^2/4);
integral(g,-10,10)
To see the curves for above:
fcurv = 1/sqrt(2*pi)*exp(-x.^2/2);
gcurv = 1/sqrt(2*pi*2)*exp(-x.^2/4);
plot(x,fcurv); hold on;
plot(x,gcurv,'r')

その他の回答 (2 件)

jenka
jenka 2013 年 4 月 24 日
Hi Wayne, yes, I tried both ways already. However, if you do sum(y) or sum(y1) to give you the are under the curve (or trapz(y)), it does not give you 1. That is why I posted here. I need the area under the curves to be equal to one. Any suggestions? Thanks

Wayne King
Wayne King 2013 年 4 月 24 日
Hi Jenka, you cannot just do sum(y), you are forgetting about the very important dx in the integral
x = -10:0.01:10;
y = normpdf(x,0,1);
y1 = normpdf(x,0,sqrt(2));
dx = mean(diff(x));
sum(y*dx)
sum(y1*dx)
  1 件のコメント
jenka
jenka 2013 年 4 月 25 日
But of course:) He-he....

サインインしてコメントする。

カテゴリ

Help Center および File ExchangeLinear and Nonlinear Regression についてさらに検索

タグ

Community Treasure Hunt

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

Start Hunting!

Translated by