Permutation DRR In Perm Nrm Btw Perm NrmA pdf with a full description of the normalization + the formulas can be found [here] The PermutationDRRInPermutationNormalizedBetweenPermutations- Normalized normalization extends the PermutationDRRIn- PermutationNormalized normalisation- reader. It reads its data through that one and then it uses the data to also apply the normalisation between permutations. So all values are transformed linearily to become 1 on average also between permutations Note that the normalisation is done per value (ngram) over all permutations. So this normalisation values all ngrams with the same weight This normalization uses temp-data. It's slow when ran for the first time (and it uses a few gb of space depending on the size of your data-set), then it becomes much faster, because it then just reads in what it calculated before Usefull for finding the ngrams that make the most difference relatively (not absolute, because all ngrams will have the average value 1, no matter the frequency differences between them) |
MenuList
