I want to compare power law distribution with log-normal using likelihood ratio test.I use power law package and use the code below with continuous data set of one variable.According to clauset et al.(2009) If this p-value is small ( p < 0.1) then it is unlikely that the observed sign is a chance result of fluctuations and the sign is a reliable indicator of which model is the better fit to the data. If p is large on the other hand, the sign is not reliable and the test does not favor either model over the other.But in case of my data, the p-value is smaller than 0.1, hence the sign is a reliable indicator of which model is the better fit.by using the below code I got 500 rows with 2 column of x and ratio. but the sign of ratio is both positive and negative. So, I don;t know using the below code, which one is the log likelihood ratio to infer from that. An screen of output is attached. the code is :
m1 = conpl$new(data) m1$setXmin(estimate_xmin(m1))
########################################################
##Exponential #
#######################################################
m2 = conlnorm$new(data) m2$setXmin(m1$getXmin()) est2 = estimate_pars(m2) m2$setPars(est2$pars)
########################################################
##Vuong's test #
########################################################
comp = compare_distributions(m1, m2)
