Monte Carlo bootstrap method for (1000 repetitions)

3 次查看（过去 30 天）

Jesper Sam 2020-9-1

0
链接

此问题的直接链接

https://ww2.mathworks.cn/matlabcentral/answers/587102-monte-carlo-bootstrap-method-for-1000-repetitions

评论： Jesper Sam 2020-9-27

hello everyone

I have global results of matrix size 360x720; but i want to apply sigificance test (Monte Carlo bootstrap method for (1000 repetitions)) for each grid poing to test The chance of coincidentally getting as high or higher ratios. please anyone can help me, where i can get this file Monte Carlo bootstrap method (code). and how can i apply this test for each grid point.

4 个评论
显示 2更早的评论隐藏 2更早的评论

Jeff Miller 2020-9-2

Sorry, I do not yet understand your dataset or computations.

Does each little triangle on your world map correspond to one cell of your 360x720 matrix? Does each little triangle represent a variance ratio?

How do you compute the variance ratio for one cell/triangle? It seems you must have many observations associated with each cell/triangle in order to compute a variance. And to compute a variance ratio for each cell/triangle you would need two variances, one for the numerator of the ratio and one for the denominator.

Bootstrapping would involve some kind of resampling of your original full dataset that went into these computations, but it is impossible to say how that might be done without knowing exactly what the original data look like.

Jesper Sam 2020-9-27

@Jeff Miller

Thankyou so much for your good reply, please i have time series of size (lat,lon,time)=360x720x36. and i need to apply Monte Carlo bootstrap method for (1000 repetitions)) for only time i.e 1x36

but i realy don't know how to use this method for significance test. second i don't have experience before use of monte carlo. therefore please if you can help me for this out

input and output of my data is same but i need output with monte carlo. can you just tell me how to use monte carlo for one time series for 1000 repetation

Thanks

请先登录，再进行评论。

请先登录，再回答此问题。