| Abstract [eng] |
Projects purpose – to describe homogeneity tests for related censored samples. It were challenges: 1. To learn how simulate Clayton‘s copula , to know what it is Clayton‘ copula; 2. To describe homogeneity tests for related censored paired samples; 3. To test homogeneity hypotesis, when data is simulated by Clayton‘s copula. And the results are: If we have censored data, with Clayton‘s parameter and ,and distributions are the same, significance level is near chosen asymptotic significance level. If we have censored data, with Clayton‘s parameter and ,and distributions are not the same, the test power for most parametres is 1 and it means that the hypotesis was rejected for all samples, and only for several test power wasn‘t 1 when the sample is n=1000. So, for my simulated samples, the homogeneity test for related censored samples, was effective. |
