Section: New Results

MonteCarlo test for two patterns of point processes on a grid

Participants : Anne Gégout-Petit, Marie Chavent, Amaury Labenne.

In order to compare two patterns of distribution of symptomatic or dead vines in a same vineyard but for two consecutive years, we have developed a Monte Carlo test. First we estimate the intensity of occurrence of disease in one of the pattern, then we simulate $n$ realizations i.i.d. of this intensity and compute the associate likelihoods in order to build an interval that cover $(1-\alpha )$ per cent of the realizations. The test reject the equality of repartition if the likelihood computed with the second pattern is not included in this interval. We have made simulations and applied this test to the repartition of esca in vineyard. This work has been presented in a national workshop on software R [46] .