| |
Methods defined here:
- __delitem__(self, i)
- Remove the descriptor number i ( x.__delitem__(i) <==> del x[i] ).
- __getitem__(self, i)
- Return a COPY of descriptor number i.
(x.__getitem__(i) <==> x[i]).
- __iadd__(self, lex)
- Add the COPYs of the Descripteurs of Lexique lex. The already
existing Descripteur numbers of self are replaced by the ones of lex.
- __init__(self, **kw)
- Keyword argument:
str -- build from string str;
alpha -- alph (with keyword str), uses alphabet alpha;
Lprop -- build from Lproportion Lprop;
fic -- build from filename fic;
fprop -- build from Lproportion file of name fprop;
- __iter__(self)
- Iterator on descriptors. Use: for element in l.
- __len__(self)
- Return the number of descriptors.
- __setitem__(self, i, x)
- If i is int : add the COPY of Descripteur x to number i;
If i is a tuple of int: build the descriptor pattern from the tuple or list of
Descripteur or string x .
- __str__(self)
- (x.__str__()<==> str(x))
- g_inter(self, a, b, v)
- Set the probability of transition between descriptors.
- is_empty(self)
- llh(self, D, **kw)
- Return the best log-likelihood on sequence D.
Keyword arguments:
deb -- first position of the computed segment of the data (default:0);
fin -- last position of the computed segment of the data
(default: the last one).
- log_exp_a_posteriori(self, D, n)
- log_likelihood(self, D, n)
- Return the list of the log-likelihoods of the data D
on the segmentations up to n classes.
- log_prob_a_posteriori(self, D, n)
- Return the list of the log-a posteriori probability of the
set of the n-partitions given the data D.
- ls_evalue(self, D, **kw)
- Return the dictionnary {tuple of descriptors
numbers:prediction of this (pattern of) descriptor on D}.
Keyword arguments:
deb -- first position of the computed segment of the data (default:0);
fin -- last position of the computed segment of the data
(default: the last one).
- ls_num(self)
- Return the list of descriptors numbers.
- met_au_net(self)
- Remove identical patterns and repetitions in patterns.
- prediction(self, D, **kw)
- Return the prediction of the best descriptor on data D.
Keyword arguments:
deb -- first position of the computed segment of the data (default:0);
fin -- last position of the computed segment of the data
(default: the last one).
- read_Lprop(self, lp)
- Build from Lproportion lp. Compute automatically
log-probabilites, or -10000 if log not possible.
- read_nf(self, nf, ns='')
- Build from filename nf and alphabet ns="".
- read_str(self, nf, ns='')
- Build from string nf and alphabet ns="".
- val_max(self, D, **kw)
- Return the maximum prediction possible on data D.
Keyword arguments:
deb -- first position of the computed segment of the data (default:0);
fin -- last position of the computed segment of the data
(default: the last one).
- windows(self, D, t, p)
- On data D, return the list of best predictions on a sliding
window of size t with slidings steps of size p.
|