Why incorrect sampling with weighted randsample?

1 次查看（过去 30 天）
Kristin
Kristin 2014-2-21
评论： the cyclist 2014-2-21
Depending on the distribution of the weights w used in randsample(population,k,true,w), the random sampling seems to be incorrect. For example, this problem arises if the weights are as weight2 below, but works correct with weight1:
%% --------------------------------------------------------- %%
values = [-5:5];
weight1 = [0.02 0.03 0.05 0.1 0.0 0.2 0.2 0.1 0.05 0.03 0.02];
weight2 = [0.02 0.00 0.05 0.1 0.35 0.28 0.2 0.0 0.00 0.00 0.00];
n = 1000000;
sample1 = randsample(values,n,true,weight1);
sample2 = randsample(values,n,true,weight2);
%% --------------------------------------------------------- %%
Has anyone else experienced this problem? Is there a workaround?
Best regards Kristin

请先登录，再回答此问题。

采纳的回答

the cyclist
the cyclist 2014-2-21
Can you please explain the evidence that makes you say that "the random sampling seems to be incorrect"?
When I run the commands
figure
hist(sample1,values)
figure
hist(sample2,values)
the results look as I would expect.
  2 个评论
Kristin
Kristin 2014-2-21
I'm so sorry, I made a simple mistake in plotting the distributions. Your reply helped me to see my mistake!
Thanks for taking your time!
/Kristin
the cyclist
the cyclist 2014-2-21
The best form of thanks is the acceptance of an answer. :-)

请先登录，再进行评论。

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Logical 的更多信息

标签

尚未输入任何标签。

Community Treasure Hunt

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

Start Hunting!

Translated by