Closed chucklesoclock closed 9 years ago
highly frequent words might have their probability adjusted to zero if theres no other words with +1 higher frequncy
that issue can be overcome by fitting the frequency of frequencies to a log-log line
test work when copied and pasted into julia, not when ran using using TextMining reload("runtests.jl")
highly frequent words might have their probability adjusted to zero if theres no other words with +1 higher frequncy
that issue can be overcome by fitting the frequency of frequencies to a log-log line
test work when copied and pasted into julia, not when ran using using TextMining reload("runtests.jl")