Hi,
I'm using markcorr for calculating the mark correlation function of an observed pattern in R,
one can assume that the observed pattern mentioned is not spatially random (according to pcf)
However, I'm not sure how to calculate the Monte Carlo (MC) envelope for significance testing here:
1) should I leave the observed pattern as it is for every MC simulation, and only switch the marks randomly at each MC simulation
or
2) should I also also create realisations of a null model (e.g. CSR) which means creating different spatial patterns for each MC simulation and then additionaly switch the marks at each MC simulation
or
3) it does not matter
I hope the question is clear,
regards,
Jan
[[alternative HTML version deleted]]