smoothing: smoothing
In gowachin/BeeMarkov:

Use the export table of viterbi function and apply two algorithms to smooth the data by two different variables

1 2	smoothing(seq, l_word_pos = 5, l_word_neg = 4, smooth_win = 10, reject_win = 1)

`seq`	export table of viterbi function
`l_word_pos`	a value for lengths of words for the model. Equal to the "order of the model + 1"
`l_word_neg`	a value fof lengths of words for the model. Equal to the "order of the model + 1"
`smooth_win`	minimal region length for a region to be kept under model + or -
`reject_win`	minimal region length for a region ambiguous to be kept under model ambiguous