KStest for unimodal and bimodal models

Hi, i would like to perform a KStest of an unimodal and bimodal models with observed data X.

Please find an example of X in attachement.

The unimodal model is a lognormal distribution defined as:

where

The bimodal model is a lognormal & lognormal PDF defined as:

where

How can i use the KStest for evaluate the goodness of fit of both models with observed data?

Is there a quick way to evaluate this with KS statistic?

T PLOCOSTE

Answer 1

[Jyotsna Talluri]

Hi,

Kstest(x) returns a test decision for the null hypothesis that the data in vector x comes from a specified standard normal distribution or it does not come from such a distribution

If the result is one the test rejects the null hypothesis at default significance level(5%)

From your data

Data=ObservedData(:,1); 
x=(Data-3.2)/0.44; %specified mean and standard deviation
r=kstest(x) ;

Similarly ,for a bimodal distribution, check the kstest for each distribution seperately

Refer to the link kstest for more information